Skip navigation

Evaluation of Proposition 36 Dept of Alcohol and Drug Programs California Health and Human Services Agency 2008.pdf

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
Evaluation of
Proposition 36:
The Substance Abuse and
Crime Prevention Act of 2000
2008 Report
Darren Urada, Ph.D., Angela Hawken, Ph.D., Bradley T. Conner, Ph.D., Elizabeth Evans,
M.A., M. Douglas Anglin, Ph.D., Joy Yang, M.P.P., Cheryl Teruya, Ph.D., Diane
Herbeck, M.A., Jia Fan, M.S., Beth Rutkowski, M.P.H., Rachel Gonzales, Ph.D.,
Richard Rawson, Ph.D., Christine Grella, Ph.D., Michael Prendergast, Ph.D., Yih-Ing
Hser, Ph.D., Jeremy Hunter, M.A. and Annie Poe, M.P.P.

Prepared for the Department of Alcohol and Drug Programs
California Health and Human Services Agency

University Of California Los Angeles

Integrated Substance Abuse Programs

TABLE OF CONTENTS
EXECUTIVE SUMMARY ......................................................................................................... 3
INTRODUCTION .................................................................................................................. 13
Preface .......................................................................................................................... 13
Chapter 1: Proposition 36 Offender Characteristics ..................................................... 17
Chapter 2: Treatment .................................................................................................... 33
SPECIAL POPULATIONS...................................................................................................... 57
Chapter 3: High-Risk and High-Cost Offenders in Proposition 36 .............................. 57
Chapter 4: The Homeless Mentally Ill in Proposition 36 ............................................. 69
PROMISING PRACTICES ..................................................................................................... 99
Chapter 5: Emerging Promising Practices in Proposition 36 ....................................... 99
Chapter 6: Employment Assistance in Proposition 36 ............................................... 105
Chapter 7: Treatment Process Improvement Methods and their Application to
Proposition 36 ............................................................................................................. 125
Chapter 8: Narcotic Treatment Programs ................................................................... 159
Chapter 9: Residential Treatment ............................................................................... 173
Chapter 10: Testing and Sanctions for Proposition 36 Probation Violations ............. 185
OUTCOMES AND PERFORMANCE ..................................................................................... 201
Chapter 11: Re-Offending and Crime Trends............................................................. 201
Chapter 12: Proposition 36 Benefit-Cost Analysis..................................................... 215
Chapter 13: Performance Monitoring & Outcome Measurement in Drug Treatment
Systems ....................................................................................................................... 231
GLOSSARY ........................................................................................................................ 265
Glossary 1: Glossary of Terms ................................................................................... 265
Glossary 2: Glossary of Abbreviations....................................................................... 266
APPENDICES .................................................................................................................... 269
Appendix A: UCLA’s Proposition 36 Stakeholder Survey ........................................ 269
Appendix B: UCLA’s Proposition 36 Treatment Program Survey ............................ 285
Appendix C: UCLA’s Proposition 36 Focus Groups Information ............................. 301
Appendix to Chapter 1................................................................................................ 305
Appendix to Chapter 2................................................................................................ 309
Appendix to Chapter 6................................................................................................ 310
Appendix to Chapter 7................................................................................................ 315

1

2

EXECUTIVE SUMMARY
This Proposition 36 (Substance Abuse Crime Prevention Act) 2008 evaluation report has
four sections. The first section describes the characteristics of clients and of treatment
utilization and outcomes in Proposition 36. The second section provides information on two
special sub-populations in Proposition 36: high risk/high cost offenders and homeless
mentally ill offenders. The third section provides preliminary evaluations of several
promising practices with potential to improve treatment program performance and client
outcomes in Proposition 36. The final section examines re-offending outcomes, presents the
findings of a cost offset analysis, then discusses additional measures that can be used to
monitor performance and outcomes in Proposition 36.
Introduction: Proposition 36 Offender Characteristics and Treatment Utilization
Offender Characteristics
A total of 50,732 offenders were referred for treatment during Proposition 36’s fifth year
(July 1, 2005-June 30, 2006). Of this number, 71.4% entered treatment.
Demographic characteristics of Proposition 36’s fifth year treatment clients were similar to
those in prior years.
•
•

•

•

Less than half (43.9%) of clients were non-Hispanic White, just over one-third
(35.9%) were Hispanic, and 13.6% were African-American.
The primary drug of use for over half of treatment clients was
methamphetamine (57.0%), followed by cocaine/crack (13.1%), marijuana
(12.5%), alcohol (8.2%), and heroin (8.0%).
The average age of Proposition 36 clients at admission was 34.8 years, and for
about half (50.4%) the Proposition 36 admission represented their first entry
into substance abuse treatment.
Most Proposition 36 treatment clients (86.8%) were sentenced to probation or
were already on probation when they committed their Proposition 36-eligible
offense. The remainder (13.2%) were on parole.

In Proposition 36’s sixth year (July 1, 2006-June 30, 2007), 48,996 offenders were referred
and 70.8% entered treatment. Recent and ongoing improvements in data collection may
have affected show rate calculations, so comparisons between these two years and prior
years must be made with caution. Show rates will be much more precisely comparable in
future years. Demographic characteristics of sixth year treatment clients were similar to
those of fifth year clients.
Treatment
Treatment placement, duration, and completion rates require passage of time to obtain
applicable data. Hence the following results apply to clients admitted in Proposition 36’s
fourth year (July 1, 2004 – June 30, 2005). Findings were very similar to those seen in prior
years.

3

•
•

•
•
•
•

Outpatient drug-free (non-narcotic replacement therapy) treatment was the
most common modality for Proposition 36 clients (84.1%).
The completion rate was 32.2% among offenders who entered treatment in
Proposition 36’s fourth year and had a final discharge on record. This is
fairly typical of results seen in studies of drug users referred to treatment by
criminal justice sources.
About half of Proposition 36 outpatient drug-free clients (45.2%) received at
least 90 days of treatment, as did 36.6% of long-term residential clients.
Treatment completion rates were lower, and treatment duration shorter, for
African-American and Hispanic groups than for other ethnic groups.
Treatment completion rates were lower, and treatment duration shorter, for
parolees relative to probationers.
Treatment completion rates were lower, and treatment duration shorter, for
opiate users compared to users of other drugs. These poorer outcomes may
be attributable to the fact that few users were treated with narcotic
replacement therapy (NRT; e.g. methadone maintenance). Opiate users in
drug free outpatient treatment had a median time in treatment of 62.5 days,
while those enrolled in NRT had a median time in treatment of 108 days.

Special Populations
High Risk and High Cost Offenders
The arrest and court costs that accumulate when offenders commit new crimes are a
significant driver of later costs under Proposition 36.
•

•

•

•

•

The typical (median) Proposition 36 offender contributes little to arrest and
court costs, while a small number of offenders contribute disproportionately
to these costs. Specifically, 25% of Proposition 36 offenders account for
80% of later crime costs.
Only 14% of those high-cost offenders who entered treatment had a
successful treatment completion compared to about one third of all
Proposition 36 clients.
Proposition 36 participant demographics were not strong predictors of later
recidivism. High-cost offenders had the same race/ethnic profile as lower
cost offenders, but were more likely to be male and, on average, about three
years younger.
A strong predictor of follow-up recidivism was the number of convictions in
the 30 months preceding an offender’s entry into Proposition 36. The perday on street arrest and conviction costs were 26 times higher for those with
5 or more prior convictions than those who entered with no prior convictions.
Identifying high-risk offenders early and managing them differently is
recommended. Possible responses might include excluding them from
Proposition 36 eligibility, incapacitating these offenders during their
participation in Proposition 36 by requiring residential treatment; or

4

•

intensively supervising these offenders while they participate in Proposition
36. Advantages and disadvantages of each are discussed.
Responding appropriately to the supervision and treatment needs of high-risk
Proposition 36 participants will be a challenge given the current limited
funding available for Proposition 36.

Homeless Mentally Ill
Previous research has shown that between 55% and 69% of individuals in the general
population who are diagnosed with an alcohol or drug use disorder have also been diagnosed
with a co-occurring mental health disorder. This population is also more likely to be
homeless. UCLA examined practices used in the assessment and treatment of mentally ill
and homeless clients in Proposition 36 and also described client characteristics and
outcomes. Findings included:
•

Approximately two-thirds of Proposition 36 county lead agencies reported
conducting a screening for mental disorders either routinely (31.25%), or in
response to elevations on typical assessment tools that indicated the presence
of a mental disorder (31.25%). The remaining lead agencies (37.5%)
reported that they did not conduct a screening for mental disorders of any
kind during the assessment process.

•

Proposition 36 drug treatment providers reported that, on average, 20.6% of
their clients were homeless and had a co-occurring mental disorder at
treatment entry. However, among drug free programs responding to UCLA’s
Treatment Program Survey, 28.6% reported employing mental health
professionals such as psychiatrists, psychologists, and/or social workers.

•

California Department of Mental Health (DMH) administrative data indicate
that Proposition 36 clients that also received mental health services in the 12
months following their Proposition 36 conviction were retained for
significantly fewer days in drug treatment and that those that were both
homeless and receiving mental health services spent the fewest days in
treatment when compared to those not homeless and not identified in DMH
administrative data. Homeless offenders eligible for Proposition 36 who
were also receiving mental health services were more likely to get arrested
for new drug, property and violent crimes than comparison groups in the 30
months following their Proposition 36 conviction, indicating that this is a
very difficult population to treat effectively.

Integrated Dual Diagnosis Treatment (IDDT) is an evidence-based practice for the treatment
of co-occurring disorders. Finding ways to build IDDT into Proposition 36’s current
treatment regimen may improve outcomes associated with the treatment of the homeless who
have co-occurring disorders.
UCLA also recommends that the state integrate Proposition 36 and Proposition 63 funding
sources to allow the creation of “Whatever It Takes” approaches to treating Proposition 36
clients who are homeless and have mental illness. This could be accomplished by awarding
Proposition 36 contracts and Proposition 63 grants to IDDT facilities.

5

Promising Practices
One goal of the current evaluation was to review a number of evidence-based strategies that
can be used to reduce no show rates into Proposition 36 treatment, to retain offenders that are
placed in treatment, and to improve program outcomes. A number of recommendations have
been generated as part of previous UCLA evaluation reports, some of which are expanded
upon in this report. These recommendations included:
•
•

•

•

•

•
•

Fund residential treatment for those with severe drug dependence and
narcotic treatment for clients with heroin or other opiate use problems.
Use practices associated with better assessment and treatment show rates,
including locating assessment units in or near the court, performing
assessments in a single visit, allowing walk-in assessments without
appointments, and incorporating procedures used in drug courts. Financial
incentives should be considered for counties and providers who institute
these or other evidence based practices or for otherwise demonstrating more
success on objective measures such as reduced time from Proposition 36
conviction to treatment entry.
Handle offenders with high rates of prior convictions differently, including
placement into residential treatment, providing more intensive supervision, or
referring to drug court.
Encourage collaboration and coordination among court, probation, parole,
and treatment providers with the goal of admitting offenders into appropriate
treatment in the shortest possible time while maintaining appropriate levels of
oversight and supervision.
Use drug testing to provide an objective basis for delivery of additional
services or for a program of graduated sanctions for offenders who are not
complying with Proposition 36 requirements.
Streamline access to state data across agencies for authorized evaluation
studies.
Continue research to address remaining issues.

Implementation of some of these recommendations has been facilitated by the Offender
Treatment Program (OTP). OTP provided $25 million in funding in 2006-2007 and $20
million in 2007-2008 to counties that apply and meet eligibility criteria. However,
Proposition 36 funding declined by $20 million in 2007-2008, so OTP funds essentially
represented a shift in source rather than new funds. Results for 2006-2007 show a 97.3%
increase in narcotic treatment program clients and an 8% increase in residential clients
among counties that requested and received funding to expand these services.
The promising practices section of this report focuses on a selected number of practices that
hold strong potential to improve Proposition 36 performance and outcomes. These include
some practices already being facilitated by OTP, such as greater use of narcotic treatment
programs, residential treatment, and drug testing and sanctions, as well as the introduction of
new practices such as employment assistance and program process improvement
6

mechanisms. UCLA recommends that funding also be made available for these practices
through OTP or more ideally from funds from a more permanent and predictable source.
Employment Assistance
Employment has long been associated with better drug treatment outcomes and this is true
for Proposition 36 clients. Significantly more of those who received employment services
also completed drug treatment successfully (51.3% vs. 38.5% in one subsample of clients).
County stakeholders also reported anecdotal success with several employment strategies in
focus group interviews. The evidence for these strategies is not strong, so it is important to
note that the following only represent potential ideas that require further study, not
recommendations. The following are examples of a few of the more innovative strategies
discussed in focus groups that could be evaluated further:
•
•

•
•

•

•

Provide job lists of “felon-friendly” employers or seasonal employers who
may be more willing to hire individuals with a criminal history.
Provide counseling to address client fears about disclosing their criminal
history to prospective employers as well as insecurities related to weak work
histories.
Provide “social events” for clients to make contacts with employed peers.
Provide a broad range of skills training and employment services at the same
location as treatment or transport clients to and from the location where such
services are offered.
Make Proposition 36 program requirements flexible enough to accommodate
the schedule of clients who are employed (e.g. night and weekend treatment
sessions).
Consider making employment a criterion for treatment completion and/or
Proposition 36 program completion.

Treatment Program Process Improvement
Methods pioneered in business settings to increase efficiency and productivity have been
applied to community-based substance abuse treatment organizations at relatively low cost
with impressive results. In 2005-2006, seven Los Angeles County treatment programs
participated in a demonstration project to apply a process improvement model developed by
the nationwide Network for the Improvement of Addiction Treatment (NIATx).
Participating programs used the recommended process improvement methods to select,
implement, and test a variety of innovative strategies including same day assessments,
increased staff contact with prospective clients, consolidated intake paperwork, client
incentives, client appointment cards, and client satisfaction surveys. Aggregate data from
the 6 outpatient programs revealed a dramatic 80% reduction in assessment no-shows and a
modest 6% increase in 30-day continuation rates.
Several conclusions and recommendations can be drawn from this pilot project:
•

These methods improved show rates and time in treatment at relatively low
cost.

7

•

•

A controlled roll-out of process improvement techniques with leadership
from the California Department of Alcohol and Drug Programs and county
lead agencies would be ideal. Participants in the pilot program reported that
guidance from the Project Director and Process Improvement Coach were
instrumental in their success, and that technical assistance with data
collection was key. Without sufficient levels of support, new participating
programs may not experience the results seen in the pilot project.
Proposition 36 funding should be allowed for these efforts and continued
where success is demonstrated.
Maintenance of sustained improvement efforts will require a permanent
coaching and technical assistance infrastructure (such as a Center for Process
Improvement) to support program staff. This will be especially important to
facilitate continuing identification and adoption of process improvement
strategies.

Narcotic Treatment Programs
Despite the evidence-based utility of Narcotic Treatment Programs (NTP) for reducing drug
use and crime among opiate addicts, criminal justice policies and anti-NTP attitudes have
hampered the use of NTP, especially for offenders. NTPs were used infrequently by
Proposition 36 offenders whose primary drug was an opioid. Several recommendations are
made based on these data and data collected from focus groups that were conducted as part
of this evaluation.
•

•

•

•
•

Educators may need to be more sensitive to ideological differences of
opinion. Significant opposition to NTP exists among some stakeholders even
after dissemination of research evidence supporting its effectiveness.
Targeted education that first collects information regarding reasons for
opposition to NTP may be more effective, but in cases where opposition is
not due to a lack of knowledge, education alone may not change this view.
While NTP may not be the appropriate treatment for every Proposition 36
participant who reports an opioid as their primary drug, it is recommended by
the National Institute on Drug Abuse and the National Academy of Science
as the treatment option of first choice.
UCLA continues to urge each county to make some form of NTP available.
Buprenorphine, which may be prescribed by authorized physicians from their
office, may be an attractive alternative NTP medication for counties that do
not currently have NTP available, are unwilling or unable to open a
methadone clinic, or are looking for innovative and cost-effective ways of
implementing NTP in their county.
Dosages of NTP medications should be closely monitored for adverse effects.
Ancillary services, including counseling, should be mandatory.

Residential Treatment
As a result of Proposition 36 there was a large increase statewide in the number of clients
presenting for drug treatment and a large increase in the number of heavy-using clients in
need of more-intensive treatment services. But due to funding constraints, and other barriers
8

to treatment expansion, the increase in demand was met largely by expanding less expensive
outpatient care. This over-reliance on outpatient treatment, particularly for severe or at-risk
offenders, affects Proposition 36 treatment and criminal justice outcomes.
•

•

•

Crime outcome differences between residential and outpatient care were
largest for clients who were heavy users of methamphetamine. This suggests
that, from a criminal justice and public safety perspective, clients who are
heavy users of methamphetamine should be prioritized for residential care.
Concerns regarding the limited use of residential treatment were raised across
stakeholder groups in focus groups and surveys. Common themes were:
concerns regarding the limited availability of residential treatment slots; the
“fail-outpatient first” approach; insufficient lengths of stay in residential care;
lack of sober-living facilities and continuing of care services; and the lack of
funding available to purchase Proposition 36 residential beds and continuing
care services.
Many stakeholders noted the importance of OTP funds to pay for Proposition
36 residential beds, and expressed concerns regarding the implications of
Proposition 36 funding cuts for the future of residential placement. Inflation
has eroded the purchasing effect of Proposition 36’s flat budget over its
initial years. In the face of this erosion and recent budget cuts, counties are
likely to reduce, rather than expand, residential treatment services. Such a
response will likely have a negative impact on Proposition 36 treatment
completion rates and criminal justice outcomes.

Drug Testing and Sanctions
Many types of sanctions are available, both by the criminal justice system and by treatment
programs. These include spending days in a jury box, intensifying treatment, and increasing
the intensity of probation supervision. Several recommendations for the use of drug testing
and sanctions in Proposition 36 are made based on data collected for this evaluation through
surveys and focus groups.
•
•

•

•

High levels of support for sanctions options within Proposition 36 exist
among key stakeholders involved in managing Proposition 36.
The basic tenets of flash incarceration programs have strong theoretical
underpinnings and are well supported in the literature. Sanctions should be
swift, certain, and consistent, and the least amount of punishment necessary
to bring about the desired behavior change should be used.
There is a small but growing body of literature on testing and jail sanctions
programs showing that swift and certain, but modest, jail sanctions can bring
about positive behavior change. These programs improved outcomes only
when probation conditions and consequences were clearly articulated to
probationers, and when violations were dealt with consistently and with
certainty. Where consistency was lacking, testing and jail sanction programs
have failed.
Expanding the conditions of Proposition 36 probation to include short jail
sanctions for non-compliance has been controversial. California Senate Bill
9

1137 was passed by the legislature in 2006 and provided discretion to judges
to give short jail stays of up to ten days to motivate treatment and probation
compliance, but this bill was opposed in court by advocacy groups and an
injunction was put into place.
Outcomes and Performance
Re-offending and Crime Trends
Trends in re-arrest rates for the Proposition 36 first year cohort (2001-2002) over a 42month follow-up period replicated those outcomes reported in previous evaluation reports at
12- and 30- month follow-up intervals.
•

Re-offending was consistently lower among Proposition 36 offenders who
completed treatment compared to offenders who did not. This effect of
participation persisted even after statistically controlling for other client
background characteristics.
• The effect of Proposition 36 as a policy on re-offending was examined by
comparing re-arrests among Proposition 36 eligible offenders in Proposition
36’s first year (Proposition 36-era offenders) to similar offenders in the preProposition 36-era. Proposition 36-era offenders were somewhat more likely
to be arrested than offenders in the pre-Proposition 36-era comparison group.
This comparison may have been affected by differences in incapacitation
under the two policies; pre-Proposition 36-era offenders were more likely to
be sentenced to jail or prison.
• Patterns of re-arrests among offenders who became eligible for Proposition
36 during the second year (2002-2003) and third year (2003-2004) were
similar to the patterns seen in offenders who became eligible during the first
year, described above. However, drug and property crime arrests were
somewhat lower among each cohort of offenders compared to the one that
came before it. This trend merits continued tracking and study to understand
its causes.
• Consistent with the comparison group differences described above, increases
in statewide drug and property arrests were somewhat greater in California
since 2001 than they were nationally. Arrests for violent crimes fell slightly
more in California than they did nationally.
Benefit Cost Analysis
• UCLA conducted three studies assessing the cost implications and benefitcost ratios of Proposition 36. Each showed that Proposition 36 yielded cost
savings to state and local governments.
• Study 1 extended the baseline and follow up periods used in UCLA’s earlier
cost report from 30 months to 42 months. Here, costs for a pre-Proposition
36-era comparison group and for all first-year Proposition 36-eligible
offenders found a net savings of $1,977 per offender (N = 61,609) over a 42
month period, yielding a benefit-cost ratio of nearly 2 to 1. In other words,
$2 was saved for every $1 invested.

10

•

•

•

Study 2 used first year Proposition 36 participants who were referred to the
program. Proposition 36 participants who completed treatment achieved a
benefit-cost ratio of approximately 4 to 1 over a 42 month period, indicating
that “completers” saved $4 for every $1 allocated.
Study 3 examined follow-up costs for succeeding year as the policy matured.
Over a 30 month follow up period, the costs for jail, probation, parole, and
treatment have remained stable from year to year. Prison costs and costs for
arrest and convictions have steadily declined over the first 3 years.
Two conclusions follow from the cost analyses: Proposition 36 substantially
reduced incarceration costs and resulted in greater cost savings for some
eligible offenders than for others.

Performance and Outcome Measures
Specific program and client measures that could potentially be used to compare performance
and outcomes in Proposition 36 are discussed along with the advantages and disadvantages
of each and suggestions for improvement.
•

•

•

Several complementary measures should be used as a package to offset the
individual weaknesses inherent in each measure. For example, one possible
combination would include treatment show rates, treatment initiation within
14 days, treatment engagement within 30 days, CalOMS outcome measures
and pre-post arrest changes. All of these measures have potential and, if all
measures were used as a set, the combination of measures would monitor
performance and outcomes at the beginning of the process (treatment show
rates, treatment initiation within 14 days), during treatment (treatment
engagement within 30 days, arrests), at treatment discharge (CalOMS
discharge outcome variables), and after treatment (arrests). Some of these
variables are easier to obtain than others, however. Data on treatment
engagement, for example, is not currently readily available. Many other
useful combinations of measures are possible.
Developing case-mix adjustments may be necessary for taking into account
differences between treatment program or county contexts (e.g. types of
clients served). Potential methods for case-mix adjustments are discussed
and a list of variables that could be considered for adjustment is included.
For many of the measures discussed it will be extremely important to collect
individual identifying information on Proposition 36 participants from all 58
counties. This would allow for tracking of outcomes among offenders who
entered Proposition 36 using state administrative databases, such as from the
Department of Justice, Department of Health Care Services, Department of
Mental Health, and other state agencies. Such data would also be critical to
fill in difficult data “blind spots” where current data limitations inhibit
tracking of all Proposition 36 participants. This county-level data is the
single most critical element required to ensure the quality of future
Proposition 36 evaluation and outcome tracking efforts. For the initial
Proposition 36 evaluation, UCLA collected such information from 10
counties, but this required individual agreements with each county, resulted
11

in 10 different sets of data with differing formats and definitions, and the
flow of data ended along with the initial evaluation. New data will be needed
to effectively track more recent cohorts of Proposition 36 clients. A
statewide effort to collect standard data from all 58 counties on a continuing
(non-expiring) basis should be led by the California Department of Alcohol
and Drug Programs.
Performance and outcome measures hold substantial promise for monitoring and improving
Proposition 36 performance and outcomes. However, if used improperly or without
addressing the significant data limitations, incentive issues, and other disadvantages
associated with each measure, inaccurate data and serious unintended consequences such as
those described in this section of the report may cause the effort to do more harm than good.
Caution and careful research is urged as measures are selected and deployed.
Conclusion
This evaluation report identifies a number of areas where improvements to Proposition 36
can be achieved and suggests strategies for achieving these improvements. However, many
of these strategies have associated implementation and maintenance costs. Due in part to the
fiscal environment faced by the State of California, insufficient Proposition 36 funding levels
have eroded stakeholders’ ability to treat and monitor Proposition 36 offenders. Moreover,
unpredictability in Proposition 36 from fiscal year to fiscal year is undermining stakeholders’
ability to engage in long term planning beyond the current year. Some of the strategies
suggested in this report are ideal because they can be implemented at a relatively low cost
given adequate leadership and participation. Others, however, will require additional
funding and funding stability to be sustainable. In these cases, successful implementation
will require prioritization on the part of state, county, and/or treatment programs. Technical
assistance and incentives from the state or county agencies tied to performance monitoring
and outcome measurement can be one useful tool to facilitate improvements, but care must
be taken to avoid unintended consequences while implementing these measures.

12

INTRODUCTION
Preface
Darren Urada, Ph.D.
In November 2000, California voters passed Proposition 36, which was enacted into law
as the Substance Abuse and Crime Prevention Act, beginning July 1, 2001.
This report has four sections. The first section describes the characteristics of clients and
treatment in Proposition 36. The second section provides information on two special
populations. The third section provides preliminary evaluations of promising practices
that may potentially improve performance and outcomes in Proposition 36. The final
section examines re-offending and cost offset outcomes and discusses the advantages and
disadvantages of other means of performance monitoring and outcome measurement.
In November 2000, California voters passed Proposition 36 (Prop 36), which was enacted
into law as the Substance Abuse and Crime Prevention Act (SACPA) of 2000. Prop 361
represents a major shift in criminal justice policy. Adults convicted of nonviolent drug
offenses in California who meet eligibility criteria can now be sentenced to probation with
substance abuse treatment instead of either probation without treatment or incarceration.
Offenders on probation or parole who commit nonviolent drug offenses or who violate drugrelated conditions of their release may also receive treatment. Levels of care may include
drug education, regular and intensive outpatient drug-free treatment, short- and long-term
residential treatment, and narcotic replacement therapy (typically methadone for clients
whose primary drug is heroin). Offenders who commit non-drug violations of
probation/parole may face termination from Prop 36. Consequences of drug violations
depend on the severity and number of such violations. The offender may be assigned to
more intensive treatment, or probation/parole may be revoked.
As part of the new law, the state was required to secure an independent statewide evaluation
of Prop 36’s effects. The California Department of Alcohol and Drug Programs (ADP)
chose the University of California, Los Angeles Integrated Substance Abuse Programs
(referred to as UCLA throughout this report) to conduct an initial evaluation of SACPA,
from 2001 to 2006. Upon completion of this evaluation, ADP contracted with UCLA to
perform a second, shorter evaluation (SACPA Evaluation II). This evaluation began on
February 23, 2007 and ended on December 31, 2007. The evaluation is focused on three
topics: 1) Promising practices and performance management, 2) Special populations, and 3)
Population/Cost-offset analysis.
Evaluation Overview
A number of states have policies that are similar to Prop 36, including Arizona (Proposition
200, 1996), Maryland (SB 194, HB 295, 2004), Hawaii (SB 1188, 2002), Washington State
1

While the term SACPA accurately refers to the name of the law, the public and various stakeholders often
know and refer to the law only as it appeared on the ballot, as Proposition 36. In recognition of this common
usage, this report adopts the commonly used term “Prop 36” in place of SACPA.

13

(SB 2338, 2002), and Kansas (HB 2309, 2003). Evaluation of these initiatives has been
either inconsistent or not funded at all (Rinaldo, & Kelly-Thomas, 2005). The goal of past
and present Prop 36 evaluation reports is to provide state and national policymakers with a
unique source of information needed to make decisions about Prop 36 in California and
similar programs elsewhere.
Data for this evaluation were collected in surveys of county stakeholders, focus groups
(semi-structured in-depth discussion) with stakeholders, observation (e.g., recording of
issues raised, perceptions noted, decisions and agreements reached) at meetings, conferences,
and other events, county records, and statewide datasets maintained by human services and
criminal justice agencies.
While the “gold standard” for program evaluation is experimental comparison in which
potential participants are randomly assigned to a program group (offered an opportunity to
participate) or a comparison group (not offered that opportunity), experimental comparison
was not feasible in the Prop 36 evaluation because randomization would have meant denying
or delaying participation by offenders legally entitled to participate in Prop 36. It was
therefore necessary to take a “quasi-experimental” approach where such comparisons were
relevant. In this approach, the comparison groups were composed of subgroups of the
people who participated, and a comparison group that was composed of people who would
have been eligible for the program if it had existed at the time of their conviction.
Organization of the Report
This draft final report is divided into four sections. The first section provides preliminary
evaluations of promising practices. The second section provides information on special
populations. The third section provides information on the Prop 36 population and
outcomes.
Introduction
Chapter 1 describes the Prop 36 “pipeline” in its fifth and sixth years, spanning July 1, 2005
to June 30, 2007. This includes the number of offenders referred to Prop 36, the number
who completed their assessment, and the number who entered treatment. Characteristics of
Prop 36 treatment clients are also described.
Chapter 2 covers the types of treatment received by Prop 36 clients, the duration of their
treatment exposure, and treatment completion in relation to offender background
characteristics.
Special Populations
Chapters 3 and 4 focus on challenges associated with two special populations: high cost
offenders and the homeless mentally ill. Information on the characteristics of these
populations and suggestions for dealing with each are included.
Promising Practices
Chapter 5 provides an overview of emerging promising practices

14

Chapters 6-10 provide information on barriers and practices related to employment, process
improvement, treatment of opiate users in narcotic treatment programs, use of residential
treatment, and use of drug testing and sanctions.
Outcomes and Performance
Chapter 11 addresses the topic of offender outcomes including re-offending (new arrests) for
Prop 36’s first, second, and third year cohorts. Outcomes are tracked in relation to an
offender’s degree of participation in Prop 36 and relative to a pre-Prop 36 era comparison
group.
Chapter 12 delivers cost benefit analyses of Prop 36 in three studies. In the first study,
offenders eligible for Prop 36 were compared with a pre-Prop 36 group of offenders to
calculate costs attributable to Prop 36 as a policy. In the second study, variations in benefitcost ratios are examined in relation to Prop 36 treatment participation. In the third study,
costs in Prop 36’s second year are compared to those in Prop 36’s first year.
Chapter 13 reviews the current state of research on performance and outcome measures in
the substance abuse field and reviews the advantages and disadvantages of number of
measures in the context of measuring county performance and outcomes.
Darren Urada, Ph.D. is the principal investigator of this evaluation. Other UCLA
researchers who had key roles in the Prop 36 evaluation include M. Douglas Anglin, Ph.D.,
Bradley T. Conner, Ph.D., Liz Evans, M.A., Jia Fan, M.S., Christine Grella, Ph.D., Angela
Hawken, Ph.D., Diane Herbeck, M.A., Yih-Ing Hser Ph.D., Jeremy Hunter, M.S., Michael
Prendergast, Ph.D., Richard Rawson, Ph.D., Cheryl Teruya, Ph.D., and Joy Yang, M.P.P.
For copies of previous Prop 36 evaluation reports, see:
http://www.uclaisap.org/prop36/html/reports.html
For information about the evaluation see:
http://www.uclaisap.org/prop36/index.html
or contact:
Darren Urada, Ph.D.
UCLA Integrated Substance
Abuse Programs
1640 S. Sepulveda Blvd., Suite 200
Los Angeles, CA 90025
Email: durada@ucla.edu

Craig Chaffee
California Department of Alcohol and
Drug Programs
1700 K Street
Sacramento, CA 95811
Tel: (916) 323-2021
Email: cchaffee@adp.ca.gov

References
Rinaldo, S., Kelly-Thomas, I. (2005). Comparing California’s Proposition 36 (SACPA)
with Similar Legislation in Other States and Jurisdictions. Berkeley, CA: The Avisa
Group. Accessed at: http://www.prop36.org/pdf/ComparisonProp36OtherStates.pdf.

15

Acknowledgments
The authors would like to thank the following individuals and organizations for their
valuable assistance:
Lily Alvarez, CADPAAC, Mike Campos, Larry Carr, Craig Chaffee, Nancy Chand,
Krista Christian, Dayna Christou, Anthony Crittenden, Priyanka Doshi, Millicent Gomes,
Amber Fitzpatrick, Michael Fitzwater, Suzette Glasner-Edwards, Dannie Hoffman, David
Huang, David Illig, Dan Johnson, George Lembi, Bryce Lowe, Judge Ana Maria Luna,
Judge Stephen Manley, Kevin Masuda, Joan Mock, Katrina Parker, Marlies Perez, Tom
Renfree, Vicki Sands, Albert Senella, Katie Shaw, Kim Teruya, Betty Viscuso, Tom
Wilson, and all of the treatment programs and county agencies that participated in UCLA
focus groups and responded to surveys.

16

Chapter 1: Proposition 36 Offender Characteristics
Darren Urada, Ph.D. and Elizabeth Evans, M.A.
A total of 50,732 offenders were referred for treatment during Prop 36’s fifth year (July
2005 through June 2006). 48,996 were referred in the sixth year (July 2006 through June
2007). Similar to previous years slightly more than 7 out of 10 referred offenders entered
treatment.
Similar to prior years, in its fifth year most Prop 36 treatment clients (73.0%) were men.
About half (43.9%) were non-Hispanic White, while 35.9% were Hispanic and 13.6%
were African-American. Their average age was 34.8 years. The primary drug of use for
over half of Prop 36’s treatment clients was methamphetamine (57.0%), followed by
cocaine/crack (13.1%), marijuana (12.5%), alcohol (8.2%), and heroin (8.0%).
Most Prop 36 offenders admitted to treatment (86.8%) were sentenced to probation or
were already on probation when they committed their Prop 36-eligible offense. The
others (13.2%) were on parole.
A large portion of Prop 36 treatment clients had never received treatment before (50.4%).
Prop 36 client characteristics have remained remarkably stable over the first five years of
operation. However, changes may occur as stakeholders respond to the identification of
areas of particular need as well as fluctuations in funding for Prop 36.
This chapter describes the “pipeline” of offenders entering Prop 36 during its fifth year.
Three steps in the pipeline are covered: referral of the offender to Prop 36, completion of the
assessment process, and entry into the treatment program to which the offender was
assigned. Show rates at assessment and treatment (i.e., the percentage who completed the
assessment process and the percentage who went on to enter treatment) in Prop 36’s fifth
year are compared to those in Prop 36’s prior years. This chapter also reports characteristics
of offenders who entered treatment during Prop 36’s fifth year.
Prop 36 Pipeline
Individuals convicted of a nonviolent drug offense, typically possession of or being under the
influence of an illicit drug, are eligible for Prop 36. As shown in Appendix 1.1, there are
some eligibility exceptions as well as differences in eligibility criteria for probationers and
parolees (Appendix 1.2).
Some offenders who are eligible for Prop 36 may decide not to participate. Those also
eligible for a “deferred entry of judgment” program1 such as PC 1000 may choose that
option because they can participate without entering a guilty plea; participation in Prop 36 is
1

Many first-time California drug offenders can avoid criminal convictions by opting for deferred entry of
judgment (DEJ) under Penal Code sections 1000-1000.4. Diversion may include education, treatment, or
rehabilitation. Entry of judgment may be deferred for a minimum of 18 months to a maximum of three years.
Although there are limitations, successfully completed diversion leads to a dismissal of the charges.

17

contingent on having been found guilty of a Prop 36-eligible offense. Moreover, depending
on local policy and practice, offenders may be eligible for both Prop 36 and drug court.
Finally, routine criminal justice processing may seem preferable to offenders who face only a
short jail sentence or disposition that they view as less onerous than the requirements of Prop
36 participation. For these reasons, it is important to assess the acceptance of Prop 36 by
eligible offenders (i.e., How many chose to participate in Prop 36 when offered that option?).
Offenders who were eligible and chose to participate in Prop 36 were ordered to complete a
treatment assessment and enter treatment. This group is known as those “referred” to Prop
36. Assessment entails a systematic review of the severity of the offender’s drug use and
other problems, a decision regarding appropriate placement in a drug treatment program, and
identification of other service needs. Upon completion of the assessment, offenders must
report promptly to the assigned treatment. Therefore referral is the first step identifiable in
the Prop 36 pipeline, completion of assessment is the second step, and treatment entry is the
third. A subsequent step, treatment completion, is discussed in Chapter 2.
Information to describe the pipeline was compiled from four sources: the SACPA Reporting
Information System (SRIS) maintained by ADP, the county stakeholder survey conducted by
UCLA in 2007, the California Alcohol and Drug Data System (CADDS), and the California
Outcomes Measurement System (CalOMS).
The first two sources were created specifically for Prop 36 monitoring and evaluation. The
third, CADDS, predates Prop 36, having been maintained by ADP since July 1991. CADDS
was modified in 2001 to require that providers indicate whether a client was referred via
Prop 36. In 2006, CalOMS replaced CADDS as ADP’s data system.
Each data source had unique value to the pipeline analysis but was subject to limitations. To
overcome these limitations, the analysis employed a mix of data taken directly from these
sources along with estimates validated across multiple sources when possible. It is important
to note that while statewide estimates are provided, the data do not allow for exact counts of
referrals or assessments for all counties.
Offenders Assessed
In the fifth year of Prop 36, an estimated 43,219 offenders, including probationers and
parolees, completed their assessment. This number was not validated by ADP. 41,925 were
assessed in the sixth year. ADP validated the sixth year numbers via telephone contacts with
the county to ensure that unique individuals were being reported and may be more accurate
than the previous year’s numbers. This may explain the higher assessment rates (88.4%) in
year six compared to year five.
Offenders Referred
According to county responses in SRIS, 50,732 offenders were referred to Proposition 36
for treatment in its fifth year and 48,996 in the sixth year (see Figure 1.1). This includes
offenders referred by the courts and by parole agents2.
2

The SRIS manual defines “referrals” as probationers and parolees sent from the court, probation department,
or parole authority.

18

Figure 1.1
Proposition 36 Offender Pipeline
(adjusted SRIS)
Referred
(Step 1)

2005-2006

2006-2007

Assessed
(Step 2)
85.2%

50,732

83.8%
Yes 43,219

Yes 36,221

No 7,513

No 6,998

85.6%
48,996

Placed in treatment
(Step 3)

Overall Show Rate:
71.4%

82.8%
Yes 41,925

Yes 34,702

No 7,071

No 7,223

Overall Show Rate:
70.8%

Offenders Entering Treatment
The estimated total of offenders placed in treatment in Prop 36’s fifth year was 36,221.3 In
the sixth year this number was 34,702.
Across Proposition 36’s first 4 years, estimated overall show rates (i.e., percentage of
offenders who were referred to Proposition 36 and went on to enter treatment) were 69.2%,
71.4%, 72.6%, and 74.9%. These previous rates may not be directly comparable to the more
recent rates (71.4%, 70.8%), however, since as noted above, in recent years ADP has made
additional efforts to validate SRIS data with county contacts. Furthermore in 2006-2007 a
new data system (CalOMS) became available to track treatment placements. All of these
factors likely affected show rates. Therefore small apparent changes in show rates in recent
years may reflect changes in data collection methods more than changes in real show rates.
In 2007-2008, it may be possible to use CalOMS as a single data source for calculating
treatment placement numbers. As recent improvements in data collection methods become
reflected in additional years, this will allow more reliable year to year comparisons to resume
in the future. For more information on computation of the show rates, see Appendix 1.3.
Prior research has shown that one-third to one-half of drug users who schedule a treatment
intake appointment (including those referred by criminal justice, other sources, and
3

The number of unique individuals in the pipeline does not precisely match the numbers that will be discussed
later from CADDS due to differing definitions. When reporting to SRIS, counties are instructed not to count
offenders who were reported in the prior reporting period. The clients in CADDS, however, may have entered
Proposition 36 treatment both during the current and past year. However, the numbers using either definition
are similar. According to CADDS, 40,358 Proposition 36 clients entered treatment during year 5, while the
pipeline estimate of clients who entered treatment in year 5 but not year 4 is 38,261. Given that the
demographic characteristics of the group generally have not changed substantially from year to year, the
statistics reported here would be very similar regardless of which definition is used.

19

themselves) actually keep their appointment (Donovan et al., 2001; Kirby et. al., 1997;
Marlowe, 2002). In a sample of drug users in Los Angeles, Hser and colleagues (1998)
found that 62% of those who asked for a treatment referral followed up on the referral they
were given. Thus, show rates seen thus far in Prop 36 compare favorably with show rates
seen in other studies of drug users referred to treatment.
No-Show Rates
State and county stakeholders have expressed interest in no-show rates (i.e., offenders who
chose Prop 36 but who did not complete an assessment or enter treatment). For a direct look
at this issue, pipeline show rates can be converted to no-show rates by subtracting from 100.
Therefore, in 2005-2006 the overall show rate of 71.4% yields a no-show rate of 100 –
71.4% = 28.6%. In 2006-2007 the no-show rate is 100-70.8% = 29.2%.
Note that no-show offenders may have failed to complete assessment or enter treatment for
various reasons. For example, these offenders may have decided to decline Prop 36
participation after initial acceptance, or they may have absconded, died, or committed crimes
or probation/parole violations that precluded further participation. To explore this issue,
UCLA included the following question on a survey of county probation stakeholders (see
Appendix A). “Among offenders who opted for Prop 36 but did not enter treatment, what
proportion would you estimate did not do so for the following reasons? (If offenders did not
enter treatment for more than one reason, percentages may add to more than 100%.)”
Representatives from 29 counties responded (see Table 1.1).
Table 1.1: Reasons for offender treatment no-shows reported by probation
respondents, and range of county responses. (n=29)
Reason
Mean %
Range
Offender was re-arrested shortly after sentencing.
15.3
1-60
Offender changed mind about participating after
11.4
0-50
learning more about the Prop 36 requirements
Offender never intended to enter treatment
30.7
1-90
Offender started using drugs again
51.4
5-96
Offender couldn’t afford fees required to enter treatment
10.0
0-75
Prop 36 requirements were incompatible with other
12.5
0-89
obligations (work schedule, for example)
Other (describe)
9.3
0-20
Two participants added the following descriptions in the “other” category: Serious medical
conditions, hospitalization, deportation, sent to CDCR on parole violation, and
absconded/warrants.
The reason that attributed for the highest proportion of no-shows was “Offender started using
drugs again,” which suggests a need to move offenders into treatment more quickly. The
second highest proportion was “Offender never intended to enter treatment,” which suggests
perceived misuse of the system. However the wide range of estimates for each reason is

20

particularly striking. These may reflect imprecision in the survey question, real differences
in reasons between counties, differences in perception, or a combination of two or more.
Given the policy implications of understanding this issue, further research to better
determine why some offenders do not enter treatment is recommended.
Characteristics of Treatment Clients
This section reports characteristics of offenders who entered treatment during Prop 36’s fifth
year. Prop 36 probation and parole referrals are shown separately so that any differences
within the Prop 36 treatment client population will be apparent. Characteristics covered in
the analysis include race/ethnicity, sex, age, primary drug, and drug problem severity.
Characteristics of clients who entered treatment during Prop 36’s fifth year but who were not
part of Prop 36 are also shown. Non-Prop 36 clients are also divided into those referred by
the criminal justice system and those entering treatment by self-referral or other non-criminal
justice referral (e.g., a healthcare provider, or employee assistance program). The purpose of
comparing treatment clients by referral source is to determine the ways in which Prop 36
clients were similar to, or different from, other clients in the state’s treatment population4.
Information on the characteristics of Prop 36 clients during its first four years was provided
in earlier reports. However, that information is also entered in figures below to allow
comparisons between client characteristics over these years.
Figure 1.2 shows the breakdown of clients entering treatment by the referral source indicated
in CADDS. In its fifth year, Prop 36 accounted for 25.9% of clients entering treatment
(22.5% were referred by probation; 3.4%, by parole). Prop 36 clients accounted for 14.8%
of all treatment clients in the law’s first year, 21.2% in the second, 22.4% in the third, and
25.9% in the fourth. Thus the share of treatment capacity taken up by Prop 36 clients has
increased across years but may be leveling out.
Figure 1.2 also shows that most of Prop 36’s fifth-year offenders (86.8%) were sentenced to
probation or were already on probation when they committed their Prop 36 eligible offense.
The others (13.2%) were parolees entering Prop 36 due to a new offense or a drug-related
parole violation. The parolee portion of the Prop 36 client population has steadily increased
over time. In the first year, 8.1% of Prop 36 treatment clients were parolees, in the second,
10.4% were parolees, in the third, 11.2% were parolees, and in the fourth, 11.6% were
parolees.

4

The CADDS admission record for each client indicates the referral source as Prop 36 (court/probation or
parole), non-Prop 36 court/criminal justice, or non-criminal justice. Clients sent from non-Prop 36
court/criminal justice may be on probation, parole, incarcerated, or participating in a non-Prop 36 diversion
program (deferred entry of judgment or drug court). Non-criminal justice clients were those referred by a
healthcare provider, employee assistance program, themselves, or other sources but not by the criminal justice
system. If a client had admissions from more than one referral source during the year, including Prop 36 and a
non-Prop 36 source, the Prop 36 admission was selected and used for these analyses.

21

Figure 1.2
Treatment Clients by Referral Source
(CADDS)
58.1

60

54.7

Percent of treatment clients

54.0

54.7

54.1

40
26.8

24.6

22.9

20

21.4

19.9

19.0

22.5

21.1

20.0

13.6

3.4

2.8

2.5

2.2

1.2

0
7/1/01 - 6/30/02
(N = 162,435)

7/1/02 - 6/30/03
(N = 166,209)

Prop. 36 probation

7/1/03 - 6/30/04
(N = 164,322)

Prop. 36 parole

7/1/04 - 6/30/05
(N = 161,535)

Criminal justice non-Prop. 36

7/1/05 - 6/30/06
(N = 155,883)

Non-criminal justice

Race/Ethnicity
The racial/ethnic composition of Prop 36 treatment clients is presented in Figure 1.3. In
Prop 36’s fifth year, almost half of Prop 36 treatment clients were non-Hispanic Whites
(43.9%). Hispanics (35.9%), African-Americans (13.6%), Asian/Pacific Islanders (2.9%),
Native Americans (1.5%), and other groups (2.1%) constituted the other half of the Prop 36
client population. Figure 1.3 also shows the racial/ethnic composition of Prop 36 clients in
the first four years. The percentage of clients who were Hispanic increased slightly each
year. Other than this, there was virtually no change across years.
F igure 1.3
Race/E thn icity of Pr op . 36 T reatmen t Cli ents
(C ADDS)

43.9
35.9
2.9
1.5
2.1

2.8
1.7
2.5

13.6

13.9

14.4
2.7
1.6
4.1

13.8
2.6
1.7
2.2

2.5
1.7
2.0

45.2
32.4

31.4
14.4

20

33.9

44.8

48.0

48.4

40
30.7

Percent of Prop. 36 treatment clients

60

0
7 /1 /0 1 - 6/ 3 0/ 0 2
( N = 2 4 ,2 8 6 )

Wh ite

His pa nic

7 /1 /0 2 - 6/ 3 0/ 0 3
(N = 3 5 ,4 0 1 )

A fric an A merica n

7 /1 /0 3 - 6/ 3 0 /0 4
(N = 3 6 ,7 7 3 )

A s ia n /P acific Is lan der

22

7 /1 /0 4 - 6 /3 0 /0 5
(N = 3 9 ,2 0 2 )

N a tive Am eric an

7 /1 / 05 - 6 /3 0 /0 6
(N = 4 0 ,3 5 8 )

Oth er

Figure 1.4 presents race/ethnicity of Prop 36 probationers and parolees separately and of
clients referred by non-Prop 36 sources in Prop 36’s fifth year. The racial/ethnic
composition of all four groups was similar.

Figure 1.4
Race/Ethnicity of Treatment Clients by Referral Source

1.6
2.3

2.0

1.5

3.2
1.7
2.3

12.7

16.5

32.9

36.6
17.0
2.5
1.8

3. 0
1.5
2.2

13 .1

20

32. 7

40

43.5

44.6

43.8
36.4

Pe rcent of treatment clients

60

44.7

(CADDS), 7/1/05 – 6/30/06
(N = 155,883)

0
Prop. 3 6 pro batio n

White

Hispanic

Prop . 36 parole

African American

C riminal jus tice
n on-Prop. 3 6

Asian/ Pa cific Is lander

N on-criminal jus tice

Native American

O ther

Sex
Clients referred to treatment by Prop 36 in its fifth year were 73.0% men and 27.0% women
(See Figure 1.5). This pattern is similar to the pattern in Prop 36’s prior years.
Figure 1.5
Sex of Prop 36 Treatment Clients
(CADDS)

Percent of Prop. 36 treatment clients

100

80

73.1

72.7

72.1

73.0

72.8

60

40
27.9

27.3

26.9

27.2

27.0

20

0
7/1/01 - 6/30/02
(N = 24,286)

7/1/02 - 6/30/03
(N = 35,401)

7/1/03 - 6/30/04
(N = 36,773)

Men

Women

23

7/1/04 - 6/30/05
(N = 39,202)

7/1/05 - 6/30/06
(N = 40,358)

Figure 1.6
Sex of Treatment Clients by Referral Source
(CADDS), 7/1/05 – 6/30/06
(N = 155,883)
100

Percent of treatment clients

80.4

80

71.9

68.2
58.7

60

41.3

40

31.8

28.1
19.6

20

0
Prop. 36 probation

Prop. 36 parole

Men

Criminal justice
non-Prop. 36

Non-criminal justice

Women

Figure 1.6 shows the sex breakdown for Prop 36 clients referred by probation and parole
and for non-Prop 36 criminal justice and non-criminal justice referrals. A majority of
treatment clients in all groups were men, but this pattern is more pronounced among
clients referred to treatment by Prop 36 and other criminal justice entities than among
non-criminal justice referrals. The pattern is most pronounced among offenders referred
to Prop 36 by parole. These results are partly a reflection of the enduring difference
between men and women in the seriousness of their criminal involvement (Blumstein et
al., 1986; Gottfredson & Hirschi, 1990).
Age
In Prop 36’s fifth year, the average (mean) age among clients referred to treatment by Prop
36 was 34.8 years. Figure 1.7 shows the distribution in age among Prop 36 clients. Over
one-fifth of Prop 36 clients (24.0%) were 25 years old or younger. Most (59.1%) were
between 26 and 45 years old. Relatively few (17.0%) were 46 years or older. These findings
closely match the findings in Prop 36’s first four years.
As shown in Figure 1.8, Prop 36 clients referred by parole were older than those referred by
probation. Moreover, clients referred from criminal justice sources other than Prop 36
included a higher percentage between 18 and 25 years old than the percentage among Prop
36 clients (43.1% vs. 25.3% of Prop 36 probation and 15.4% of Prop 36 parole). Finally,
non-criminal justice referrals include more clients in the oldest age bracket. Because crime
is less prevalent in older-age cohorts (Gottfredson & Hirschi, 1990; Hirschi & Gottfredson,
1983), it is not unusual that non-criminal justice referrals include a higher percentage of
older clients.

24

Figure 1.7
Age of Prop 36 Treatment Clients
(CADDS)

1 7.0

28.3
3 0.8

24. 0

32.2

28.0

23.6

33.4

15.8

12.6

13.1

20

14.3

2 2.7

23.2

29.2

34.1

30.0

30.2

35.0

40

22.0

Percent of Prop. 36 treatment clients

60

0
7/1/01 - 6/30/02
(N = 24,286)

7/1/02 - 6/30/03
(N = 35,401)

7/1/03 - 6/30/04
(N = 36,773)

18-25

26-35

7/1/04 - 6/30/05
(N = 39,202)

36-4 5

7/1/05 - 6/30/06
(N = 40,358)

46 +

F ig u re 1 .8
A ge o f T rea tm en t C lien ts b y R eferra l S o u rce
(C A D D S ), 7 /1 /0 5 – 6 /3 0 /06
(N = 1 5 5,8 83 )

Percent of treatment clients

60

4 3 .1

40

3 5 .3
2 7 .7

3 2 .2

3 0 .1

2 9 .9

2 5 .3

20

2 3 .8
1 6 .9

1 5 .4

2 6 .0
2 1 .9

2 1 .4

2 2 .3

1 7 .2
1 1 .6

0
Prop. 36 proba tion

Prop. 36 parole

1 8 -2 5

C rim in a l ju s tice
n o n -Prop. 36

2 6 -3 5

3 6 -4 5

No n -crim in al ju s tice

46+

Primary Drug
According to client self-report, as depicted in Figure 1.9, methamphetamine was the most
common primary drug used by Prop 36 clients in the fifth year (57.0%), followed by
25

cocaine/crack (13.1%), marijuana (12.5%), alcohol (8.2%), and heroin (8.0%). These figures
are largely unchanged from Prop 36’s earlier years, except that the proportion of clients who
reported methamphetamine as their primary drug has increased nearly every year. In
addition to the primary drug, the majority of Prop 36 clients (64.6%) also reported using at
least one other drug.
Primary drug by referral source is presented in Figure 1.9. As was true in Prop 36’s earlier
years, methamphetamine use was more common in Prop 36 clients than in the other two
client groups. Moreover, within the Prop 36 treatment population, heroin use was more
common among parolees (11.8%) than among probationers (7.4%). Heroin use was more
prevalent among non-criminal justice clients (24.4%) than among criminal justice clients,
possibly because heroin users may, on their own initiative (self-referral), seek methadone
treatment to avoid the symptoms of heroin withdrawal. CADDS reporting requirements may
also increase the prevalence of reported heroin use relative to other drugs. Specifically,
private as well as publicly funded providers are required to report methadone treatment
admissions to CADDS, whereas only publicly funded providers are required to report
admissions to other types of treatment programs.
F igure 1.9
Prim ary Dru g A mong T reatmen t Cli ents by Referral S ou rce

41.4

56.5

57.1

24.7

24.4
9.6

14.2

20.1

8.8

1.5

0.9

1.3

5.2

8.2

8.8
11.8

8.2
1.2

7.4

20

13.9

23.2

25.7

40

13.0
13.1

Percent of treatment clients

60

(CADDS) , 7/1/05 – 6/30/06
(N = 155,883)

0
Pro p . 3 6 p ro ba t io n

Meth am ph eta mine

P rop . 3 6 p ar ol e

C o cain e/crack

C ri min a l ju st ic e
n o n- Pro p . 3 6

M arijua na

Heroin

N o n - cri min al j us t ic e

Alco ho l

Oth er

In Figure 1.10, alcohol was the self-reported primary drug for 8.2% of the Prop 36 group,
even though Prop 36 targets offenders with illicit drug offenses. Heavy drinking is quite
common among people who use illicit drugs. Figure 1.11 shows the secondary drug
recorded in CADDS for Prop 36 clients whose self-reported primary drug was alcohol. The
distribution of secondary drug mirrors the distribution for primary drug. Methamphetamine
was the most common secondary drug (32.9%). Cocaine (17.3%) and marijuana (21.3%)
were also prevalent. No secondary drug was shown for 24.5% of Prop 36 clients whose
primary drug was alcohol. These findings for Prop 36’s fifth year closely parallel those for
the prior years.

26

F ig ur e 1. 10
P r ima r y D r ug Am ong P r op 36 Tr e a tm en t C lien ts
55.3

52.7

53.0

50.2

1.2

1.4

8.0
8.2

8.6
8.8

13.1
12.5

13.7
12.7
1.7

1.5

1.7

13.2
12.1
10.2
9.8

14.5
11.7
10.6
10.6

20

14.3
12.2
9.6
9.4

40

0
7/ 1/ 01 - 6/ 30/ 02
(N = 24, 286)

7/ 1/0 2 - 6/ 30/0 3
(N = 35, 401)

M e th a m p h et a m in e

7/ 1/03 - 6/30 /04
(N = 36, 773)

C o ca in e/c ra c k

7/1/ 04 - 6 /30/ 05
(N = 3 9,20 2)

M a rij u an a

H e ro in

7 /1/ 05 - 6/ 30/ 06
(N = 40 ,358 )

A lco h o l

O t he r

Figure 1.11
Secondary Drug when Alcohol is Primary Drug
Among Prop 36 Treatment Clients

24.5

32.9

2.9
1.1

2.5
1.3

17.3
21.3

22.5

19.9
3.5
1.9

3.1
1.9

18.8
20.8

33.1
16.4

21.2
20.4

21.5
21.5

18.5
3.9
2.3

20

19.5
23.0

32.4

35.3

40

34.0

(CADDS)

60
Percent of Prop. 36 treatment clients

Percent of Prop. 36 treatment clients

60

57.0

( CA D DS )

0
7/1/01 - 6 /30/02
(N = 2, 579)

7/1/02 - 6/30/03
(N = 3,488)

Methamphetamine

7/1/03 - 6/30/04
(N = 3 ,470)

Cocaine/cra ck

Marijuana

27

7/1/04 - 6/30/05
(N = 3,425)

Heroin

Other

7/1/05 - 6/30/06
(N = 3,297)

None

Clients with alcohol as their primary drug and no secondary drug on record may have
reported a secondary drug that was not entered into CADDS, or may have failed to report a
secondary drug despite having one. In any case, clients reporting alcohol as a primary drug
with no secondary drug constituted less than 2% of the Prop 36 fifth year client population
and had no substantial impact on the patterns reported below.
Drug Problem Severity
UCLA analyzed three indicators of drug problem severity: years of primary drug use,
frequency of recent drug use, and prior treatment experience.
Figure 1.12 shows a split distribution of drug use histories among Prop 36 treatment clients.
About one-fifth of Prop 36’s clients in each year (22.4% in the fifth year) reported first use
of their primary drug within the last five years. One-quarter (25.5% in the fifth year)
reported primary drug histories extending longer than 20 years.
Figure 1.12
Years Since First Use of Prim ary Drug
A m ong Prop 36 Treatm ent Clients
(C AD D S)

22.9
20.9
16.6
15.5
24.1

21.9
20.6
17.7
14.8
25.0

22.4
20.0
17.9
14.2
25.5

20

20.7
22.0
17.3
15.9
23.9

40

20.7
21.5
17.7
16.1
23.7

Percent of Prop. 36 treatment clients

60

7 /1 /0 1 - 6 /3 0 /0 2
(N = 2 4 ,2 8 6 )

7 /1 /0 2 - 6 /3 0 /0 3
(N = 3 5 ,4 0 1 )

7 /1 /0 3 - 6 /3 0 /0 4
(N = 3 6 ,7 7 3 )

7 /1 /0 4 - 6 /3 0 /0 5
(N = 3 9 ,2 0 2 )

7 /1 /0 5 - 6 /3 0 /0 6
(N = 4 0 ,3 5 8 )

0
0 -5

6 - 10

11 - 15

16 - 20

2 1+

Figure 1.13 shows years since first use of primary drug by referral source for the fifth year
population. Prop 36 parolees reported longer primary drug histories than Prop 36
probationers and non-Prop 36 criminal justice referrals. About one-third (30.6%) of Prop 36
parolees reported having used their primary drug for more than 20 years.

28

Figure 1.13
Years Since First Use of Primary Drug
Among Treatment Clients by Referral Source
(CADDS), 7/1/05 – 6/30/06
(N = 155,883)

34.6
27.9

11.0

14.3

12.2

20.9
11.0

18.6

21.8

11.6

17.4

18.6

24.8
13.7

20

17.3

20.2

24.1

30.6

35.0

40

14.5

Percent of treatment clients

60

0
Prop. 36 probation

Prop. 36 parole

0 -5

6 - 10

Criminal justice
non-Prop. 36

11 - 15

16 - 20

Non-criminal justice

21+

Frequency of primary drug use by Prop 36 clients in the month prior to treatment admission
is shown in Figure 1.14. About one-third (40.9%) of fifth year Prop 36 clients reported no
primary drug use in the past month, possibly because they were entering treatment directly
from being incarcerated5 or had ceased use due to probation or parole oversight. Previous
Prop 36 evaluation reports also reported this pattern.
F ig u r e 1 .1 4
F r e q u e n c y o f P r i m a r y D r u g U se in P a s t M o n t h
A m o n g P r o p 3 6 T r e a t m e n t C li e n t s
( C ADD S)

40.9

37.1

14.8

13.4
9.2

10.0

11.8

16.2

24.9

27.5
10.1

12.2

21.7

34.7
15.4
9.6

14.9

12.4

9.9

20

12.1

27.0

28.8

34.1

35.2

40

15.5

Percen t of Prop . 36 treatment clients

60

0
7 /1 /0 1 - 6 / 3 0 /0 2

7 /1 /0 2 - 6 /3 0 /0 3

7 /1 /0 3 - 6 /3 0 /0 4

7 /1 /0 4 - 6 / 3 0 /0 5

( N = 2 4 ,2 8 6 )

( N = 3 5 ,4 0 1 )

(N = 3 6 ,7 7 3 )

( N = 3 9 ,2 0 2 )

No n e

1 - 3 ti m es /m o n th

1 - 2 t im e s / w ee k

5

3 - 6 t im e s /w e e k

7 /1 /0 5 - 6 /3 0 /0 6
( N = 4 0 ,3 5 8 )

D a ily

In a prior offender survey (see 2004 report), about 60% of offenders who reported no drug use in the month
before treatment entry had been in jail (55.8%) or inpatient healthcare (3.3%).

29

As shown in Figure 1.15, Prop 36 and non-Prop 36 criminal justice clients were more likely
to report no primary drug use in the past month compared to non-criminal justice clients.
Non-criminal justice clients conversely were far more likely to report daily drug use in the
past month. This divergence may have arisen because of the reasons listed above.

Figure 1.15
Frequency of Primary Drug Use in Past M onth
Among Treatm ent Clients by Referral Source
(CAD DS), 7/1/05 – 6/30/06
(N = 155,883)

60

Percent of treatment clients

48.0
42.3
38.3

40
31.7
26.1
22.4

21.0

18.6

18.5

20

14.6

14.2

13.2
9.3

13.312.213.9

13.5

11.6

9.1

8.3

0
Pro p. 36 probation

N one

Prop. 36 parole

1 - 3 times /month

C riminal jus tic e
non-Prop. 36

1 - 2 times /w ee k

N on-criminal jus tice

3 - 6 times/w eek

D aily

Figure 1.16
Number of Prior Treatment Admissions
Among Prop 36 Treatment Clients
50.4

49.2

46.2

48.8
7/1/01 - 6/30/02
(N = 24,286)
0

1

7/1/02 - 6/30/03
(N = 35,401)
2

3

4

30

27.9

29.1

7/1/03 - 6/30/04
(N = 36,773)
5

5.1
2.1
1.2
0.6
0.4
0.2
1.0

11.2

4.8
2.1
1.1
0.6
0.3
0.2
0.9

4.7
1.8
1.1
0.5
0.2
0.2
0.8

5.1
2.1
1.1
0.6
0.3
0.3
0.8

11.7

3.9
1.7
0.9
0.4
0.2
0.1
0.7

0

13.0

20

11.7

26.5

29.7

30.5

40

9.9

Percent of Prop. 36 treatment clients

55.2

(CADDS)
60

7/1/04 - 6/30/05
(N = 39,202)
6

7

7/1/05 - 6/30/06
(N = 40,358)
8

9+

Figure 1.16 shows the number of self-reported prior treatment admissions among Prop 36
clients. In its fifth year, slightly more than half of Prop 36’s clients (50.4%) reported no
prior experience in drug treatment. The portion of such clients decreased each year during
the first three years of Prop 36, but increased slightly during the fourth and fifth years.
Figure 1.17 compares treatment experience among clients from all referral sources. Slightly
more than half of the non-criminal justice referrals (52.1%) reported no prior treatment, a
finding very similar to that for Prop 36 referrals on probation as well as parole. Over half of
the non-Prop 36 criminal justice referrals (60.1%) reported no prior treatment.
F igu re 1.17
N u m b er of P rior T rea tm en t A d m ission s
A m o n g T reatm en t C lien ts b y R eferral S ou rce
60.1

(C A D D S ), 7 /1 /0 5 – 6 /3 0 /06
(N = 1 5 5,8 83 )
52.1

46.8

51.0

40

0

P ro p . 3 6 p ro batio n

0

1

P ro p. 3 6 p aro le

2

3

C rim ina l ju stice
n o n -P ro p . 36

4

5.8
3.4
2.4
1.5
0.8
0.6
3.9

10.2

19.2
8.7
3.8
1.8
1.0
0.5
0.3
0.2
0.7

4.9
2.0
1.1
0.6
0.4
0.2
1.0

6.3
2.7
1.7
0.9
0.4
0.3
1.2

11.0

12.7

20

23.0

27.0

28.0

Percent of treatment clients

60

5

6

N o n -crim ina l justice

7

8

9+

Conclusion
Show rates were similar to those estimated for previous Proposition 36 years. Similar to
Proposition 36’s earlier years, in its fifth year, most Proposition 36 treatment clients (73.0%)
were men; about half (43.9%) were non-Hispanic White, while 35.9% were Hispanic and
13.6% were African-American; the average age was 34.8 years; the primary drug of use for
over half of Proposition 36’s treatment clients was methamphetamine (57.0%), followed by
cocaine/crack (13.1%), marijuana (12.5%), alcohol (8.2%), and heroin (8.0%).
Most Proposition 36 offenders admitted to treatment (86.8%) were sentenced to probation or
were already on probation when they committed their Proposition 36-eligible offense. The
others (13.2%) were on parole. A large portion of Proposition 36 treatment clients had never
received treatment before (50.4%).
References
Blumstein, A., Cohen, J., Roth, J., & Visher, C. (1986). Criminal careers and “career
criminals.” Washington, DC: National Academy Press.

31

Donovan, D.M., Rosengren, D.B., Downey, L., Cox, G.B., & Sloan, K.L. (2001).
Attrition prevention with individuals awaiting publicly funded drug treatment.
Addiction, 96, 1149-1160.
Gottfredson, M.R., & Hirschi, T. (1990). A general theory of crime. Stanford, CA:
Stanford University Press.
Hirschi, T., & Gottfredson, M. (1983). Age and the explanation of crime. American
Journal of Sociology, 89, 552-584.
Hser, Y.-I., Maglione, M., Polinsky, M.L., & Anglin, M.D. (1998). Predicting drug
treatment entry among treatment-seeking individuals. Journal of Substance Abuse
Treatment, 15, 213-220.
Kirby, K.C., Marlowe, D.B., Lamb, R.J., & Platt, J.J. (1997). Behavioral treatments of
cocaine addiction: Assessing patient needs and improving treatment entry and
outcome. Journal of Drug Issues, 27, 417-429.
Marlowe, D.B. (2002). Effective strategies for intervening with drug abusing offenders.
Villanova Law Review, 47, 989-1025.

32

Chapter 2: Treatment
Darren Urada, Ph.D. and Elizabeth Evans, M.A.
Treatment placement, duration, and completion rates in Prop 36’s most recent year of
operation were very similar to patterns seen in prior years.
Outpatient drug-free (non-narcotic replacement therapy) treatment was the most common
modality for Prop 36 clients (84.1%), followed by long-term residential treatment
(11.5%).
Methadone maintenance, methadone detoxification, non-methadone
detoxification, and short-term residential treatment were rarely used in Prop 36.
Treatment placement in Prop 36’s fifth year was very similar to placement in its first four
years.
Treatment completion among Prop 36 offenders thus far is typical of drug users referred
to treatment by criminal justice. The completion rate was 32.2% among offenders who
entered treatment in Prop 36’s fourth year and had a final discharge on record.
Treatment completion rates were lower, and treatment duration shorter, for African
Americans and Hispanics than for Whites, Asians and Pacific Islanders, and Native
Americans. These findings signal the importance of addressing the possible
disproportionate impact of limited treatment capacity, assessment procedures, and
treatment protocols across racial/ethnic groups.
Clients with no prior experience in treatment may find it particularly difficult to conform
to unfamiliar requirements such as open acknowledgement of their drug problem and
self-disclosure in groups. Despite the potential difficulties, first-time clients did as well
in treatment as clients who had been in treatment before.
Methamphetamine users were similar to the overall Prop 36 population in treatment
duration and completion.
Treatment duration was shorter and completion rates lower for heroin users than for users
of other drugs. In each Prop 36 year thus far, few heroin users were treated with
methadone detoxification or maintenance.
Treatment completion was lower, and duration shorter, for parolees than for probationers
in Prop 36.
This chapter reproduces and updates analyses presented in the Evaluation of the Substance
Abuse and Crime Prevention Act Final Report (UCLA ISAP, 2007). The chapter consists of
three sections dealing with treatment placement, treatment completion, and treatment
duration. While the portion of this chapter dealing with treatment placement focuses on Prop
36’s fifth year, the portion dealing with treatment completion and duration focus on Prop
36’s fourth year so as to provide time for clients to be discharged from treatment.

33

First, the chapter reports the treatment modalities Prop 36 clients were placed in during the
fifth year. For comparison, treatment placement in Prop 36’s first four years is also
summarized.
Second, as noted, the chapter reports results from analyses of treatment completion and
duration among Prop 36’s fourth year clients. The focus is on the first four years of Prop 36
because data are not yet available to determine how Prop 36’s fifth year population will fare
after entering treatment. Treatment completion among Prop 36’s fourth year clients is
examined and compared to completion in Prop 36’s first, second, and third years. Then
characteristics of fourth year clients who completed treatment are reported. These
characteristics include, for example, race/ethnicity, sex, and primary drug.
Third, the chapter offers findings on treatment duration. Like the findings on completion,
findings on treatment duration in Prop 36’s fourth year are examined in relation to client
characteristics and compared to findings from Prop 36’s earlier years. CADDS was the data
source for these analyses.
Research on drug treatment effectiveness has shown that treatment completion and time in
treatment are associated with favorable post-treatment outcomes such as abstinence from
drug use, reductions in drug-related problems, and improved psychosocial functioning
(Anglin & Hser, 1990; DeLeon, 1991; Hubbard et al., 1989, 1997; Simpson, 1979; Simpson
et al., 1997; TOPPS II Interstate Cooperative Study Group, 2003). Thus, the performance of
Prop 36 offenders on these two indicators of treatment performance, treatment completion
and time in treatment, serves as a useful indicator of the likelihood of post-treatment success.
The analysis of these treatment measures, however, does not tell the whole story. Prop 36
clients must not only attend treatment but also must comply with other requirements set by
the court and probation/parole. Their obligations in Prop 36 are not fully met even if they do
complete treatment. However, limited statewide data is available on these final completion
specifics.
Treatment Placement
While not enough time had passed to conduct treatment completion and duration analyses for
the fifth year cohort at the time of this analysis, admission data are available for treatment
placement. Accordingly, this section refers to clients who entered treatment in Prop 36’s
fifth year.
CADDS data were analyzed to determine the percentage of Prop 36 offenders entering each
treatment modality. As shown in Figure 2.1, outpatient drug-free (non-NTP) was the initial
treatment placement for most offenders (84.1%). Long-term residential treatment (planned
duration exceeding 30 days) was the second most common placement (11.5%). This pattern
was the same regardless of the client’s primary drug (see Figure 11.2). Treatment placement
in Prop 36’s fourth year was very similar to placement in the first three years.
Methadone maintenance, methadone detoxification, non-methadone detoxification, and
short-term residential treatment were rarely used in Prop 36. Methadone maintenance and
detoxification are effective in treating heroin dependence (American Methadone Treatment
Association, Inc., 2004; Mathias, 1997; National Institute on Drug Abuse, 1999; National

34

Institutes of Health Consensus Conference, 1998). Thus it is notable that few heroin or other
opiate users in Prop 36’s fourth year (15.1%) were treated with methadone detoxification or
maintenance. Comparable data for Prop 36’s first three years were 9.9%, 12.7%, and 12.9%
respectively. The increase in methadone treatment in the fourth year was primarily
attributable to an increase in the use of methadone detoxification, which rose from 2.7% in
the third year to 6.0% in the fourth year. Most heroin and other opiate users were placed in
outpatient drug-free programs, which do not provide medication to alleviate the withdrawal
symptoms associated with heroin dependence.
F ig u r e 2 .1
P r o p 3 6 T r e a tm e n t C lie n ts b y M o d a lity
(C A D D S ), 7 /1 /0 5 – 6 /3 0 /0 6
(N = 4 0 ,3 5 8 )
Percent of Prop. 36 treatment clients

100
8 4 .1

80

60

40

20

1 1 .5
2 .2

0 .9

0 .5

0 .7

D e to x

R e s ide ntia l
< 3 0 da ys

M e th a d o ne
de to x

M e tha do ne
m a inte na n c e

0
O utp a tie nt
D rug F re e

R e s ide ntia l
> 3 0 days

Figure 2.2
Primary Drug by M odality Among Prop 36 Treatment Clients
(CAD DS), 7/1/05 – 6/30/06
(N = 40,358)
85.4

93.4
81.0

83.1

86.3

60.1

80

60

O utpatie nt

Alcohol
(N = 3 ,2 97 )

R e side ntia l
> 3 0 da ys

Cocaine/Crack
(N = 5 ,2 9 2)

D e tox

R esidential
< 3 0 da ys

35

Heroin/Other Opiates
(N = 3 ,23 3)

Marijuana
(N = 5 ,04 9)

M etha done de to x

2.4
0.2
0.0
0.0

12.0
0.4
0.6
0.0
0.0

5.7

1.2
6.0
9.1

14.0
9.6
1.9
0.7
0.0
0.0

13.0

Methamphetamine
(N = 2 3,0 1 7 )

2.7
1.1
0.0
0.0

0

1.5
1.0
0.0
0.0

20

16.4

40

11.1

Percent of Prop. 36 treatment clients

100

Other
(N = 4 6 5 )

M etha do ne ma intena nc e

Treatment Completion
Results discussed in this section apply to clients admitted to treatment during Prop 36’s
fourth year, July 1, 2004 - June 30, 2005.
Comparative Completion Rates
For a standard of comparison against which to judge Prop 36 completion rates, this chapter
summarizes findings on treatment completion from other large-scale studies of drug
treatment. In addition, completion rates for Prop 36 clients are compared to those for nonProp 36 criminal justice clients and non-criminal justice clients1 who received treatment
during the same timeframe. Finally, information on drug court completion rates is provided.
In national studies of drug treatment effectiveness, completion rates have ranged from 35%
to 60% (Substance Abuse and Mental Health Services Administration, 2002; TOPPS II
Interstate Cooperative Study Group, 2003). Treatment completion rates have also been
reported in two large-scale studies examining drug treatment effectiveness in the state of
California. The completion rate was 32% in CALDATA, fielded in the early 1990’s
(Gerstein et al., 1994). More recently, the CalTOP study (Hser et al., 2003) found that 41%
of clients with a discharge on record (excluding clients whose discharge indicated a transfer
for additional treatment) had completed treatment.
Nationally, drug court graduation rates range from 31% to 73% and average about 50%
(Belenko, 2001; Latessa et al., 2002; Logan et al., 2004; Rempel et al., 2003). In California,
graduation rates of 36% (Belenko, 2001) and 55% (California ADP, 2005) have been
reported. However it should be noted that eligibility criteria can affect drug court completion
rates. Prop 36 is open to all offenders who meet eligibility criteria, while drug courts
typically have greater discretion to determine which offenders participate.
In the United Kingdom, a community sentence for offenders who misuse drugs known as the
Drug Treatment and Testing Order was introduced in 1998. The Order requires offenders to
submit to regular drug testing, attend an intensive treatment program, and have their progress
reviewed regularly by the courts. In 2003, 28% of Drug Treatment and Testing Orders
“completed in full or terminated early for good progress.” (Bourn, 2004). An evaluator has
suggested that results would be improved if the implementation of the Order more closely
followed the model of U.S. drug courts (Bean, 2002).
Measuring Treatment Completion
To allow time for clients to participate in and be discharged from treatment, and to allow for
lag in data entry, analyses of treatment completion and duration focus on Prop 36’s fourth
year, July 1, 2004 - June 30, 2005.
In CADDS, a client’s status at discharge is noted by the treatment provider on the client’s
discharge record. There are four possible statuses at discharge: completed treatment, did not
complete treatment but made satisfactory progress, did not complete treatment and did not
1

The CADDS record for each incoming client indicates the referral source as Proposition 36 (court/probation
or parole), non-Proposition 36 court/criminal justice, or non-criminal justice. Clients sent from non-Proposition
36 court/criminal justice were generally on probation, on parole, incarcerated, or were otherwise participating
in a non-Proposition 36 diversion program. Non-criminal justice clients were referred by healthcare providers,
employee assistance programs, themselves, or other sources.

36

make satisfactory progress, and transferred to another treatment provider. The most rigorous
criterion for success is the treatment completion rate among clients with a final discharge on
record other than a referral/transfer.2 This is the measure employed in discussions of
completion below. For more on methods used to define a treatment episode, and analysis of
this measure’s sensitivity to assumptions about missing data, see Appendix 2.
Clients who did not complete treatment may also have been doing well. Clients leaving
treatment early may have found a job that required them to be at work during treatment
hours, moved to a location farther away from the treatment provider, taken on competing
responsibilities such as childcare, or lost their means of transportation. The purpose of the
“satisfactory progress” criterion is to enable providers to enter a discharge status that reflects
the opinion that a client was in recovery services long enough to have made significant
progress toward achieving the goals set forth in his/her recovery plan. This chapter also
reports the percentage of clients who did not complete treatment but made satisfactory
progress. However, it is important to emphasize that Prop 36 requires completion of
treatment. While clients who made satisfactory progress may have benefited from treatment,
they were out of compliance with the treatment requirement if they did not complete
treatment and were still subject to disqualification from Prop36 by the court.
Prop 36 Treatment Completion
As shown in Figure 2.3, 32.2% of Prop 36’s fourth year clients completed treatment. The
completion rates in Prop 36’s first, second, and third years were 34.4%, 34.3%, and 32.0%.
Prop 36’s adjusted completion rates in all four years were somewhat lower than the adjusted
rates for non-Prop 36 criminal justice clients and slightly higher than the adjusted rate for
non-criminal justice clients.
Figure 2.3 also shows clients who did not complete treatment but were making satisfactory
progress. Among Prop 36 clients, 8.1% met criteria for satisfactory progress. The adjusted
rates for non-Prop 36 criminal justice clients (11.8%) and non-criminal justice clients
(15.1%) were higher. Overall, 40.3% of Prop 36’s fourth year clients either completed
treatment or made satisfactory progress. Non-Prop 36 criminal justice clients and noncriminal justice clients had rates of 49.8% and 50.1% on this overall indicator of treatment
performance. Findings for first, second, and third-year clients were similar.
Figure 2.4 shows variability in treatment completion rates across counties. In each of Prop
36’s first four years, completion rates were between 26% and 50% in most counties. Further
research is needed to investigate why these variations occur, and whether the adoption of
practices from counties with higher rates would result in improved completion rates in
counties that reported lower rates. Variation in county completion rates may also result from
different mixes of treatment modalities, different populations, and variations in the definition
of treatment completion between counties. Standardization of the definition of treatment
completion across the state would allow for more accurate interpretation of these completion
rates and variations.
2

CADDS instructions define a treatment completer: “This participant has successfully completed his/her
recovery plan and has met the major goals set forth in that plan. The participant is not being referred or
transferred to any other alcohol or drug program.”

37

Figure 2.3
Discharge Status by Referral Source
(CADDS)
60

Percent of clients

7/1/01 – 6/30/02

40

12.1

7/1/02 – 6/30/03

36.0

34.3

34.2

15.1

9.9

7.3

34.4

7/1/04 – 6/30/05

11.8

10.3

14.7

8.0

20

7/1/03 – 6/30/04

15.0

13.8

7.4

37.5

8.1

38.0

36.9
32.0

30.0

35.0

32.2

30.6

0
Prop. 3 6
Criminal Non-criminal
P rop. 3 6
Criminal Non-criminal
P rop. 3 6
Criminal Non-criminal
Prop. 3 6
Criminal Non-criminal
(N = 1 8 ,8 4 6 ) jus tice nonjus tice
(N = 28 ,7 4 9) jus tice nonjus tice
(N = 30 ,2 4 6) jus tice nonjus tice
(N = 3 1 ,6 0 5 ) jus tice nonjus ti ce
Prop. 36
(N = 7 9 ,6 8 6 )
P rop. 3 6
(N = 7 7 ,8 66 )
Prop. 3 6
(N = 7 8 ,2 98 )
Prop. 36
(N = 7 5 ,3 4 2 )
(N = 4 1 ,81 1 )
(N = 3 8,6 1 5 )
(N = 3 6,2 7 6 )
(N = 3 2 ,00 5 )

Comple ted treatment

Made satisfactory progress

F igu re 2.4
C ou n ty V ariation in C om p letion R ates
for P rop 36 O ffen d ers
(C A D D S )

Number of counties

60

40

39

38

36

35

20
10

8

7

11

9

12
8

1

1

0

7 /1 /01 - 6/3 0/02
(N = 54)

7 /1 /0 2 - 6/3 0/0 3
(N = 56)

7/1/0 3 - 6/30 /04
(N = 5 5)

1

1

0

0 - 25%

26 - 50%

51 - 75%

7/1/0 4 - 6/30 /0 5
(N = 53 )

76+%

N ote: A nalysis for each year exc luded co unties w here no P roposition 36 c lients had a d isc harge on
record.

Characteristics and Treatment Completion
To analyze characteristics of clients who completed treatment, UCLA employed the most
rigorous criterion for success, namely a discharge record showing “completed treatment.”
UCLA conducted an analysis to see whether Prop 36 client characteristics associated with
treatment completion when taken one at a time (e.g. age, race, etc.) maintained an
association with completion when all characteristics were tested together. Findings reported
here were confirmed in that analysis.

38

As shown in Figure 2.5, Whites (35.4%) had the highest rates of treatment completion in
Prop 36’s fourth year. Asian-Americans and Pacific Islanders (34.1%), Native Americans
(33.5%), Hispanics (30.1%), and African-Americans (25.9%) followed. Patterns of
racial/ethnic differences in Prop 36 generally did not parallel patterns in non-Prop 36 groups.
Among criminal justice non-Prop 36 referrals, Asian-Americans and Pacific Islanders had
the highest treatment completion rate (43.8%), followed by Whites (41.1%), Native
Americans (37.6%), Hispanics (36.8%), and African Americans (36.8%). Among noncriminal justice referrals, Whites had the highest treatment completion rates (39.1%),
followed by African Americans (34.7%), Native Americans (34.1%), Asian-Americans and
Pacific Islanders (32.1%), and Hispanics (29.3%).

Figure 2.5
Treatment Completion Among Clients by Race/Ethnicity
(CADDS), 7/1/04 – 6/30/05
(N = 138,952)

30.9

39.5
33.1

34.1

33.5
37 .6

43.8
3 2.1

34.1

34.7

30.1

25.9

29.3

3 0.1

36.8

41.1
39.1

40

3 5.4

Percent of treatment clients

60

20

0
Whit e
N = 1 4,242 13,451 35,082

Hispanic
10,8 61 11 ,702 23,586

Prop. 36

African Ame rican
4,308 4,5 09 1 2,011

Asian/Pacific Islander
896 1,08 0 1,502

Criminal justice non-Prop. 36

Native Am erican
546 538 1,290

O the r
752 724 1 ,862

Non-criminal justice

Treatment completion rates for men and women are shown in Figure 2.6. Women in Prop 36
had slightly higher completion rates (33.3%) than men (31.8%), as has been the trend in each
of Prop 36’s first four years. Completion rates were more similar between men and women
in the criminal justice non-Prop 36 group, but were more dissimilar in the non-criminal
justice group.

39

Figure 2.6
Treatment Completion Among Clients by Sex
(CADDS), 7/1/04 – 6/30/05
(N = 138,952)

Percent of treatment clients

60

37.7

40

38.7

36.4
33.3

31.8

32.9

20

0
Me n
(N = 23,200 22,856 44,734)

Prop. 36

Women
(N = 8,405 9,149 30,608)

Criminal justice non-Prop. 36

Non-criminal justice

A positive association between age and treatment completion is apparent in Figure 2.7. The
completion rate for Prop 36 clients in the youngest age range (25 years and younger) was
28.5%. Rates climbed to a maximum of 36.7% in the oldest age range (46 years and older).
This same stair-step pattern is apparent for the two non-Prop 36 groups as well. Older drug
users may be more likely to see the value of completing treatment given the accumulation of
problems arising from their drug use over time.
Figure 2.7
Treatment Completion Among Clients by Age
(CADDS), 7/1/04 – 6/30/05
(N = 138,952)

Percent of treatment clients

60

43.2

40

38.4

36.0
28.5

38.8
33.9

30.6

31.3

39.5

37.1

36.7

33.6

20

0
25 or younger
(N =7,640 13,929 21,521)

Prop. 36

26 - 35
(N = 8,976 7,324 18,050)

36 - 45
(N = 10,084 7,166 20,757)

Criminal justice non-Prop. 36

40

46+
(N = 4,905 3,586 15,014)

Non-criminal justice

Figure 2.8 shows completion rates by primary drug. Findings are most relevant for the four
drugs commonly used by Prop 36 clients. Heroin users in Prop 36 had the lowest completion
rates (26.6%)3 . This was also true in both non-Prop 36 groups. Notably, methamphetamine
users completed treatment at rates similar to users of most other drugs.

F ig u re 2.8
Tre atm en t C om p letion A m on g C lie n ts b y P rim ary D ru g

36.7
36.9

38.9

46.8
36.1
29.8
24.8

26.6

30.3

29.2

35.7

37.8

33.3

30.2

37.4
35.2

33.4

Per cent of tr eatme nt c lients

40

45.1

(C AD DS ), 7/1/04 – 6/ 30/ 05
(N = 138,952)

60

20

0
M et h a m ph e t am in e
N = 1 7 ,50 5 12 ,2 94 2 0, 22 8

Co ca in e /cr ack
4, 24 0 4 ,2 40 7 ,5 77

P r op. 3 6

M ar ij u a n a
4 ,0 08 7 ,1 88 9 ,67 1

H ero i n /ot h e r o pi a te s
2 ,6 63 2 ,33 9 17 ,5 56

C ri mina l j ust ic e non -P r op. 36

Al co h ol
2 ,7 69 6 ,67 1 19 ,1 60

Othe r
4 17 4 22 1 ,12 9

N o n- cr imi nal j us tic e

The association between years since first use of primary drug and treatment completion (see
Figure 2.9) mirrors that between age and treatment completion due to the relationship
between age and years of use. The completion rate for Prop 36 clients with the fewest years
since first use of their primary drug (no more than five) was 30.0%. Clients with at least 21
years of use had the highest completion rate (34.8%). The two non-Prop 36 groups showed
the same pattern.
Figure 2.10 shows treatment completion rates by frequency of primary drug use in the month
prior to intake. The treatment completion rate was highest among Prop 36 clients who
reported no use at all in the past month (38.3%), perhaps because they were less likely to
experience craving/withdrawal symptoms while in treatment or because prior-month
abstinence, whether voluntary or imposed by circumstance (e.g., being in jail), was
indicative of greater motivation to stop using or of less access to drugs. Completion was
lower among all Prop 36 clients who reported any use of their primary drug in the month
prior to intake. The trend toward slightly higher completion rates among clients who
reported daily use is in part due to the higher prevalence of residential treatment among this
population.
3

See discussion earlier in this chapter on the relatively low use of NTP for heroin users in Prop 36. Completion
rates for heroin users in non-criminal justice settings are not directly comparable since they were far more
likely to enter methadone maintenance programs (see Chapter 8). “Completion” is not a meaningful measure in
such settings because the program goal in these cases is often indefinite maintenance.

41

Figure 2.9
Treatment Completion Among Clients
by Years Since First Use of Primary Drug
(CADDS), 7/1/04 – 6/30/05
(N = 138,952)

Pe rcent of treatment clients

60

37.2

40

3 7.0
31.7

30.0

31 .4

31.9

33.3

31.0

4 0.6 39.5

38.6

37.0

35.4

33.5

34 .8

20

0
0 -5
(N = 6,956 11,180 19,993)

6 - 10
(N = 6,538 6,115 11,609

Prop. 3 6

11 - 15
(N = 5,632 4,424 9,584)

16 - 20
(N = 4,664 3,572 9,200

Criminal justice non-Pro p. 36

21+
(N = 7,815 6,714 24,956)

Non-criminal justice

Figure 2.10
Treatment Completion Among Clients
by Frequency of Primary Drug Use in Past 30 Days

32.1

30.4

30.5

33.1

27.3

26.9

31.9

34 .5

3 5.4
28.0

33.2

37.0

40

38.3

43.8

(CADDS), 7/1/04 – 6/30/05
(N = 138,952)

33.9

Percent of treatment clients

60

20

0
N one
(N = 11 ,4 6 5 7 ,9 3 9 3 7, 43 1 )

Prop. 36

1 - 3 tim es/ month
(N = 5 ,1 7 4 4 ,3 69 7 ,6 6 6 )

1 - 2 ti mes /we ek
(N = 3 ,8 0 4 1 3 ,7 03 1 4 ,3 4 5 )

Criminal justice non-Prop. 36

42

3 - 6 time s/ we ek
(N = 3 ,1 47 2 ,9 5 6 6 ,2 5 2 )

D ai ly
(N = 7 ,7 5 3 2 ,7 4 0 9 ,1 15 )

Non-criminal justice

Treatment completion rates were similar for Prop 36 clients with and without prior
experience in treatment (32.7% and 31.5%, respectively). This was true in the non-Prop 36
groups as well (See Figure 2.11).
As shown in Figure 2.12, Prop 36 clients on probation (33.3%) had a somewhat higher
completion rate than clients on parole (24.4%). Parolees were older, reported using drugs for
longer periods, and were more likely to report daily use and use heroin. However, even after
controlling for these factors a difference remains. By definition, parolees are supervised by a
different system (parole rather than probation) and they tend to have more serious criminal
histories than do probationers. Further study of the parole subpopulation and parole
procedures associated with success and failure is warranted. The figure does not include
non-Prop 36 groups because CADDS data on non-Prop 36 criminal justice referrals do not
distinguish between probation and parole and this distinction is not applicable to noncriminal justice referrals.

Figure 2.11
Treatment Completion Among Clients
by Prior Treatment Experience
(CADDS), 7/1/04 – 6/30/05
(N = 138,952)

Percent of treatment clients

60

38.2

37.6

40

35.0

34.9

32.7

31.5

20

0
Prior admission
(N = 15,539 13,057 36,589)

Prop. 36

No prior admission
(N = 15,634 18,606 38,373)

Criminal justice non-Prop. 36

Non-criminal justice

Treatment Duration
Treatment Duration among Clients Who Completed Treatment
Similar to the findings on completion, findings on treatment duration in Prop 36’s fourth
year were examined in relation to client characteristics and compared to findings from Prop
36’s first three years.
Clients were classified as receiving as outpatient or residential depending on their initial
placement. Most Prop 36 clients (93.9%) were ultimately discharged from the same
treatment modality as the one they were initially placed in. For clients whose treatment
episode included two or more segments, either in the same type of treatment or in different
43

Figure 2.12
Treatment Completion Among Prop 36 Clients
Referred by Probation and Parole
(CADDS), 7/1/04 – 6/30/05
(N = 31,605)

Percent of Prop. 36 treatment clients

60

40
33.3
24.4

20

0
Probation
(N = 27,713)

Parole
(N = 3,892)

Median days of treatment

Figure 2.13
Median Length of Stay in Treatment
Among Treatment Completers by Modality
300
270
240
210
180
150
120
90
60
30
0

(CADDS), 7/1/04 – 6/30/05
(N = 48,665)

176

171

143146
110
90 90 90
57
30 30
15

13 10 6
Outpatient Drug
Free
Prop. 36
(N = 10,169)

Residential
> 30 days

Detox

Residential
< 30 days

Criminal justice non-Prop. 36
(N = 12,159)

20 20 20

Methadone detox

Methadone
maintenance

Non-criminal justice
(N = 26,337)

Note: Findings for methadone maintenance may be unreliable for Prop. 36 and Criminal justice
non-Prop. 36 clients because numbers were small (Prop. 36 n=35, Criminal justice n=15).

44

types, the calculation of treatment duration covered their total time in treatment from first
intake to last discharge. Therefore the times in treatment reported below may include time
spent in a modality different from the first (e.g. a client may have spent time in residential
treatment after initially entering outpatient treatment). These charts only include the subset
of clients with a discharge on record of “completed treatment”. Across the state, median
time to treatment completion was 176 days for Prop 36 clients in outpatient drug-free
treatment and 90 days for those in long-term residential treatment (See Figure 2.13). In Prop
36’s first four years, median times to treatment completion were longer for outpatient drugfree and similar for long-term residential treatment.
Among clients referred from criminal justice sources other than Prop 36, demographicadjusted median duration for completers was 143 days in outpatient drug-free treatment and
90 days in long-term residential treatment. Non-criminal justice clients who completed
treatment typically spent an adjusted median of 146 days in outpatient drug-free treatment or
90 days in long-term residential treatment. Prop 36 clients who completed outpatient drugfree programs had somewhat longer stays than non-Prop 36 outpatient drug-free clients.
Residential stays were the same across groups.
Counties varied widely on the number of days that Prop 36 clients were in treatment prior to
being discharged with a successful completion. Figure 2.14 shows the distribution of
counties for outpatient drug-free treatment. While the median duration was over 300 days in
6 counties, the median was no more than 200 days in 34 counties4. Figure 2.15 shows the
distribution of counties for long-term residential treatment. The median was less than 200
days in most counties. However, the median was over 200 days in five counties5.
A period of at least 90 days is widely cited as the minimum threshold for beneficial treatment
(Hubbard et al., 1997; Simpson et al., 1997, 1999, 2002; TOPPS II Interstate Cooperative
Study Group, 2003). The typical fourth year Prop 36 client who completed residential
treatment reached this threshold, and the typical outpatient client in Prop 36 exceeded it (see
above). The 90-day threshold remains a useful benchmark for evaluating exposure to
treatment among Prop 36 clients, regardless of how much longer they may have stayed,
whether they completed treatment, or how well they fared. This analysis reports the
percentage of fourth year Prop 36 clients who remained in outpatient drug-free treatment or
long-term residential treatment for at least 90 days and who had a discharge record. To
account for clients who did not receive at least 90 days of treatment, the analysis was
expanded to show the percentage spending at least 30 days and at least 60 days in each
treatment modality. Findings are compared across years and examined in relation to client
demographic characteristics. For clarity of presentation, detailed information on treatment
duration among non-Prop 36 clients is omitted.
4

Three counties were excluded because the number of clients who completed outpatient treatment was too
small to support a reliable estimate of treatment duration. Since modality is defined by the client’s first
admission but duration attempts to capture the entire course of treatment, the durations reported here may
include time spent in other modalities that the client transferred to, including residential treatment.
5
Eight counties were excluded because the number of clients who completed residential treatment was too
small to support a reliable estimate of treatment duration. Since modality is defined by first admission but
duration attempts to capture the entire course of treatment, the durations reported here may include time spent
in other modalities the client transferred to following the initial admission, including outpatient treatment.

45

Figure 2.14
County Variation in Median Length of Stay
Among Outpatient Treatment Completers
(CADDS), 7/1/04 – 6/30/05
(N = 54)
34

35

Number of counties

30
25
20
15

13

10
6

5
1

0
1 - 100

101 - 200

201 - 300

301 +

Number of days

Note: In 3 counties, the number of outpatient treatment completers was too low for a reliable
estimate of length of stay. Yuba and Sutter County results are combined.

Figure 2.15
County Variation in Median Length of Stay
Among Residential Treatment Completers
(CADDS), 7/1/04 – 6/30/05
(N = 49)

30

Number of counties

25

25

19

20
15
10

4

5

1

0
1 - 100

101 - 200

201 - 300

300 +

Number of days
Note: In 8 counties, the number of residential treatment completers was too low for a reliable estimate
of length of stay. Yuba and Sutter County results are combined.

46

Treatment Duration among All Clients
Most Prop 36 clients (72.9%) who entered outpatient drug-free programs were there for at
least 30 days (see Figure 2.16). Among long-term residential clients, 70.7% received at least
30 days of treatment. The 60-day rates were 56.7% in outpatient drug-free treatment and
51.1% in long-term residential treatment. Finally, about half of Prop 36 outpatient drug-free
clients (45.2%) received at least 90 days of treatment, as did 36.6% of long-term residential
clients.

Figure 2.16
T reatm ent D uration A m ong Prop 36 C lients by M odality
(CA DD S), 7/1/04 – 6/30/05
(N = 30,007)

Percent of Prop. 36 treatment clients

10 0

80

72.9

70.7
56.7

60

51.1
45.2
36.6

40

20

0
30 days

O utpatient (N = 26,393 )

60 days

90 days

Lo ng -te rm res ide ntial (N = 3 ,61 4)

Characteristics and Treatment Duration
UCLA examined treatment duration in relation to the following background characteristics
of Prop 36 clients: race/ethnicity, sex, age, primary drug, years of primary drug use, recent
frequency of use, and referral source (probation or parole). Clients in outpatient drug-free
treatment and long-term residential treatment were combined. Figure 2.17 shows treatment
duration by race/ethnicity of Prop 36 clients.
The percentage of Prop 36 clients who reached 90 days was slightly lower among AfricanAmericans, Hispanics, and Native Americans than among Whites, Asian-Americans, and
Pacific Islanders.
Figure 2.18 shows treatment duration for Prop 36 clients by sex. Men and women in Prop 36
had similar patterns of duration at 30, 60, and 90 days.

47

F ig u r e 2 .1 7
T r ea tm en t D u r a tio n A m o n g P r o p 3 6 C lien ts
b y R a c e /E th n ic ity
(C A D D S ), 7 /1 /0 4 – 6 /3 0 /0 5
(N = 3 1 ,6 0 5 )

Percent of Prop. 36 treatment clients

100

80

60

7 5 .5

7 3 .0

7 0 .1
5 6 .8

7 1 .3

6 8 .0
5 8 .0
5 3 .2

4 4 .4

5 0 .7

4 2 .0

5 4 .4

5 2 .7

4 8 .0
4 2 .3

4 1 .2

3 9 .5

40

7 0 .1

20

0
W h ite
(N = 1 2 ,2 4 2 )

H is pan ic
(N = 1 0 ,8 6 1 )

A fric an A m e ric an
(N = 4 ,3 0 8 )

30 days

60 days

A s ian/P ac ific
Is land e r
(N = 8 9 6 )

N at iv e A m e ric an
(N = 5 4 6 )

O th e r
(N = 7 5 2 )

90 days

Figure 2.18
Treatment Duration Among Prop 36 Clients by Sex
(CADDS), 7/1/04 – 6/30/05
(N = 31,605)

Percent of Prop. 36 treatment clients

80

71.7

71.1

60

55.0

54.5

43.6

42.7

40

20

0
Men
(N = 23,20 0)

Wome n
(N = 8,405)

30 days

60 days

90 days

Treatment duration by age is shown in Figure 2.19. At all three intervals, duration rates were
slightly higher among older Prop 36 clients.
Treatment duration by primary drug is shown in Figure 2.20. Clients who entered treatment
with a primary drug of methamphetamine, cocaine/crack, and marijuana had similar duration
patterns at 30, 60, and 90 days. Clients whose primary drug at admission was heroin or
another opiate were somewhat less likely to reach 90 days. However it is important to note
48

Figure 2.19
Treatment Duration Among Prop 36 Clients by Age
(CADDS), 7/1/04 – 6/30/05
(N = 31,605)
Percent of Prop. 36 treatment clients

80

60

46.2

43.9

42.3

40.3

58.6

55.6

53.9

51.6

73.2

72.4

70.8

69.2

40

20

0
18 - 25
(N = 7,640)

26 - 35
(N = 8,976)

30 days

36 - 45
(N = 10,084)

60 days

46+
(N = 4,905)

90 days

Figure 2.20
Treatment Duration Among Prop 36 Clients
by Primary Drug
(CADDS), 7/1/04 – 6/30/05
(N = 31,605)

Percent of Prop. 36 treatment clients

100

80

60

72.3

73.4

70.4

53.2
43.2

72.7
57.2

54.5

61.3

44.7

43.5

57.1
44.6

40

72.2

45.3

54.9
45.8

34.3

20

0
Methamphetamine
(N = 17,505)

Cocaine/crack
(N = 4,240)

Marijuana
(N = 4,008)

30 days

Heroin/other opiates
(N = 2,663)

60 days

49

90 days

Alcohol
(N = 2,769)

Other
(N = 417)

that those enrolled in drug free outpatient treatment had a median time in treatment of only
62.5 days, while opiate users enrolled in narcotic replacement therapy (e.g. methadone
maintenance) had a substantially higher median time in treatment (108 days). See Chapter 8
for further discussion of narcotic replacement therapy.
As shown in Figure 2.21, there was no relationship between years of primary drug use and
treatment duration among Prop 36 clients.
F ig u r e 2 .2 1
T r e a t m e n t D u r a tio n A m o n g P r o p 3 6 C lie n ts
b y Y e a r s o f P r im a r y D r u g U s e
(C A D D S ) , 7 /1 /0 4 – 6 /3 0 /0 5
( N = 3 1 ,6 0 5 )
Percent of Prop. 36 treatment clients

100

80

7 0 .9 7 1 .1

6 9 .5

60

7 2 .7

7 1 .2

4 2 .9

4 1 .9

5 6 .4

5 5 .7

5 4 .2

5 2 .8

7 2 .4

4 3 .8

4 3 .7

4 2 .2

40

20

0
0 -5
(N = 6 ,9 5 6 )

6 - 10
(N = 6 ,5 3 8 )

30 days

11 - 15
(N = 5 ,6 3 2 )

60 days

16 - 20
(N = 4 ,6 6 4 )

21+
(N = 5 ,6 6 1 )

90 days

Figure 2.22 shows treatment duration by frequency of primary drug use in the 30 days before
treatment entry. The percentage of Prop 36 clients in treatment at each interval declined as
F ig u r e 2 .2 2
T r e a tm e n t D u r a tio n A m o n g P r o p 3 6 C lie n ts
b y F r e q u e n c y o f P r im a r y D r u g U se in P a s t 3 0 D a y s
(C A D D S ), 7 /1 /0 4 – 6 /3 0 /0 5
(N = 3 1 ,6 0 5 )

Percent of Prop. 36 treatment clients

100

80

7 8 .1
7 1 .3

6 7 .9

6 3 .6

60

5 1 .6

5 4 .3

6 7 .3

5 0 .6
4 2 .6

6 4 .7
4 8 .8

3 9 .9

40

4 5 .9
3 6 .4

3 4 .2

20

0
N o ne
(N = 1 1 ,4 6 5 )

1 - 3 time
(N = 5 ,1 7 4 )

30 days

1 - 2 time s /we e k
(N = 3 ,8 0 4 )

60 days

50

90 days

3 - 6 tim e s /we e k
(N = 3 ,1 4 7 )

D a ily
(N = 7 ,7 5 3 )

Figure 2.23 shows treatment duration for Prop 36 clients with and without treatment
experience. There was no relationship between treatment duration and prior treatment
experience.

Figure 2.23
Treatment Duration Among Prop 36 Clients
by Prior Treatment Experience
(CADDS), 7/1/04 – 6/30/05
(N = 31,173)

Percent of Prop. 36 treatment clients

100
80

71.4

71.0

60

54.8

54.2
43.0

42.6

40
20
0
Prior admissions
(N = 15,539)

30 days

No prior admissions
(N = 15,634)

60 days

90 days

Figure 2.24 shows duration patterns separately for Prop 36 clients on probation and those on
parole. Parolees were less likely to be in treatment at each interval.

Percent of Proposition 36 treatment clients

Figure 2.24
Treatment Duration Among Prop 36 Clients
by Referral Source
(CADDS), 7/1/04 – 6/30/05
(N = 31,173)

100

80

72.2
64.7

60

55.9
45.9

44.3

40

32.7

20

0
Probatio n
(N = 27,7 13)

30 days

Parole
(N = 3,89 2)

60 days

51

90 days

Discussion and Conclusions
Treatment placement, duration, and completion rates in Prop 36’s most recent year of
operation were very similar to patterns seen in prior years.
Most treatment clients in each of Prop 36’s first five years (84.1% in its fourth year) were
placed in outpatient drug-free treatment.
Prop 36 clients appeared to be faring about as well as others receiving treatment in the same
timeframe. The rate of successful treatment completion was 32.2% among offenders who
entered treatment in Prop 36’s fourth year and had a final discharge on record. These
findings, which were similar in Prop 36’s first three years, are typical of drug users referred
to treatment by criminal justice.
A total of 40.3% of Prop 36’s third year clients either completed treatment or were making
satisfactory progress when discharged. Treatment completion and satisfactory progress are
good signs, but it is important to note that successful completion of Prop 36 also requires
compliance with the conditions of probation/parole supervision.
In Prop 36, treatment completion rates were lower and 90-day treatment duration less
common for African-Americans and Hispanics, than for Whites, Asian-Americans, Pacific
Islanders, and Native Americans. The same was true in Prop 36’s earlier years. Disparities
in completion rates may reflect entrenched societal conditions. Nevertheless, these
disparities are cause for concern. It may be important to explore opportunities to improve
cultural competence in assessment and treatment of Prop 36 clients. Cultural competence
reflects an “awareness of cultural differences and the development of skills to work in
multicultural situations” (Campbell et al., 2002, page 110; see also Betancourt et al., 2003)
and is believed to have a positive impact on health service utilization, sustained participation,
satisfaction with services, and outcomes (Campbell et al., 2002; Paniagua, 1994; Resnicow
& Braithwaite, 2001; Smedley et al., 2003). Alternatives for promoting cultural competence
include racial/ethnic matching between staff and clients, offering clients the opportunity to
choose a counselor of the same race/ethnicity, offering single-race group counseling sessions
or self-help support groups, hiring personnel who are bilingual, and training staff in crosscultural awareness and skills.
Methamphetamine users were similar to the overall Prop 36 population in treatment
completion and duration in each Prop 36 year analyzed. Concern has been raised regarding
the treatment system’s ability to meet the clinical challenges presented by methamphetamine
users (e.g., poor engagement in treatment, severe paranoia, severe and protracted dysphoria,
and high relapse rates; Rawson et al., 2002). Findings suggest that treatment providers in
Prop 36 have responded to the challenges presented by methamphetamine users.
In Prop 36’s fourth year, treatment completion was lower, and duration shorter, for users of
heroin than for users of other drugs. Chapter 5 of this report provides in-depth information
on Prop 36 treatment of heroin users.
Clients with no prior experience in treatment may find it particularly difficult to conform to
unfamiliar requirements such as open acknowledgement of their drug problem and self-

52

disclosure in groups. Despite the potential difficulties, first-time clients were as likely to
complete treatment as clients who have been in the treatment system previously.
Completion rates were lower, and treatment duration shorter, for parolees than for
probationers in both Prop 36 years. This finding suggests a need to evaluate and implement
improvements for parolees. Possibilities for consideration include increased supervision,
increased use of dedicated Prop 36 agents, and closer collaboration between parole agents,
county agencies, and treatment providers.

References
American Methadone Treatment Association, Inc. Fact sheet: Why methadone treatment
works. American Association for the Treatment of Opioid Dependence. Accessed
at: www.aatod.org/factsheet1_print.htm.
Anglin, M.D., & Hser, Y.-I. (1990). Treatment of drug abuse. In Tonry, M. & Wilson,
Q. (Eds.), Drugs and crime. Chicago: The University of Chicago Press.
Bean, P. (2002). Drugs and Crime. Portland, OR: Willan Publishing.
Bourn, J. (2004). The Drug Treatment and Testing Order: early lessons. Accessed at:
from: http://www.nao.org.uk/publications/nao_reports/03-04/0304366.pdf.
Belenko, S. (2001). Research on drug courts: a critical review. New York: The
National Center on Addiction and Substance Abuse at Columbia University.
Accessed at: www.drugpolicy.org/docUploads/2001drugcourts.pdf.
Betancourt, J.R., Green, A.R., Carrillo, J.E., & Ananeh-Firempong, O. (2003). Defining
cultural competence: A practical framework for addressing racial/ethnic disparities in
health and health care. Public Health Reports, 118, 293-118.
California Department of Alcohol and Drug Programs (2005). Comprehensive drug
court implementation act of 1999: Interim report to the legislature. Sacramento,
CA.
Campbell, C.I., & Alexander, J.A. (2002). Culturally competent treatment practices and
ancillary service use in outpatient substance abuse treatment. Journal of Substance
Abuse Treatment, 22, 109-119.
De Leon, G. (1991). Retention in drug-free therapeutic communities. In Pickens, R.W.,
Leukefeld, C.G., & C.R. Schuster (Eds.), Improving drug abuse treatment. NIDA
research monograph 106 (DHHS Publication No. ADM 91-1754). Rockville, MD:
National Institute on Drug Abuse. Accessed at:
www.nida.nih.gov/pdf/monographs/download106.html.
Gerstein, D., Dean, R., Johnson, R., Foote, M., Suter, N., Jack, K., Merker, G., Turner,
S., Bailey, R., Malloy, K., Williams, E., Harwood, H. & Fountain, D. (1994).
Evaluating Recovery Services: The California drug and alcohol treatment
assessment (CALDATA) methodology report. Sacramento, CA: California
Department of Alcohol and Drug Programs.

53

Hser, Y.-I., Evans, E., Teruya, C., Hardy, M., Ettner, S., Urada, D., Huang, D., Picazo,
R., Shen, H., Hseih, J., Anglin, M. D. (2003). The California Treatment Outcome
Project (CalTOP) final report. Los Angeles, CA: UCLA Integrated Substance
Abuse Programs. Accessed at: www.uclaisap.org/caltop/index.htm.
Hubbard, R.L., Marsden, M.E., Rachal, J.V., Harwood, H.J., Cavanaugh, E.R., &
Ginzburg, H.M. (1989). Drug abuse treatment: A national study of effectiveness.
Chapel Hill, North Carolina: University of North Carolina Press.
Hubbard, R.L., Craddock, S.G., Flynn, P.M., Anderson, J., & Etheridge, R.M. (1997).
Overview of 1-year follow-up outcomes in the Drug Abuse Treatment Outcome
Study (DATOS). Psychology of Addictive Behaviors, 11, 261-278.
Latessa, E.J., Shaffer, D.K., & Lowenkamp, C. (2002). Outcome evaluation of Ohio’s
drug court efforts. Center for Criminal Justice Research, Division of Criminal
Justice, University of Cincinnati.
Logan, T.K., Hoyt, W.H., McCollister, K.E., French, M.T., Leukefeld, C., & Minton, L.
(2004). Economic evaluation of drug court: Methodology, results, and policy
implications. Evaluation and Program Planning, 27, 381-396.
Mathias, R. (1997). NIH panel calls for expanded methadone treatment for heroin
addiction. NIDA Notes, 12. Accessed at:
www.drugabuse.gov/NIDA_Notes/NNVol12N6/NIHPanel.html.
National Institute on Drug Abuse (1999). Principles of drug addiction treatment (NIH
Publication No. 99-4180). Washington DC: National Institutes of Health. Accessed
at: www.nida.nih.gov/PODAT/PODATindex.html.
National Institutes of Health Consensus Conference (1998). Effective medical treatment
of opiate addiction. Journal of the American Medical Association, 280, 1936-1943.
Retrieved on April 9, 2004, from
www.odp.od.nih.gov/consensus/cons/108/108_intro.htm.Retrieved on April 9, 2004,
from www.odp.od.nih.gov/consensus/cons/108/108_intro.htm.
Paniagua, F.A. (1994). Assessing and treating culturally diverse clients: A practical
guide. Thousand Oaks, CA: Sage Publications.
Rempel, M., Fox-Kralstein, D., Cissner, A., Cohen, R., Labriola, M., Farole, D., Bader,
A., & Magnani, M. (2003). The New York State adult drug court evaluation:
Policies, participants and impacts. New York: New York State Unified Court
System and the U.S. Bureau of Justice Assistance, Center for Court Innovation.
Retrieved on April 9, 2004, from
www.courtinnovation.org/pdf/drug_court_eval_exec_sum.pdf.
Resnicow, K., & Braithwaite, R.L. (2001). Cultural sensitivity in public health. In
Braithwaite, R.L. & Taylor, S.E. (Eds.), Health Issues in the Black Community. San
Francisco: Jossey-Bass Publishers.
Smedley, B.D., Stith, A.Y., & Nelson, A.R. (2003). Unequal treatment: Confronting
racial and ethnic disparities in health care. Washington, DC: The National
Academies Press.

54

Substance Abuse and Mental Health Services Administration, Office of Applied Studies
(2002). Treatment Episode Data Set (TEDS): 1992-2000. National admissions to
substance abuse treatment services (Publication No. SMA 02-3727). Rockville,
MD: Department of Health and Human Services. Accessed at:
www.dasis.samhsa.gov/teds00/TEDS_2k_index.htm.
Simpson, D.D. (1979). The relation of time spent in drug abuse treatment to posttreatment outcome. American Journal of Psychiatry, 136, 1449-1453.
Simpson, D., Joe, G.W., & Broome, K.M. (2002). A national 5-year follow-up of
treatment outcomes for cocaine dependence. Archives of General Psychiatry, 59,
538-544.
Simpson, D., Joe, G.W., & Brown, B.S. (1997). Treatment retention and follow-up
outcomes in the Drug Abuse Treatment Outcome Study (DATOS). Psychology of
Addictive Behaviors, 11, 294-307.
Simpson, D., Joe, G.W., Fletcher, B.W., Hubbard, R.L., & Anglin, M.D. (1999). A
national evaluation of treatment outcomes for cocaine dependence. Archives of
General Psychiatry, 57, 507-514.
TOPPS II Interstate Cooperative Study Group (2003). Drug treatment completion and
post-discharge employment in the TOPPS-II Interstate Cooperative Study. Journal
of Substance Abuse Treatment, 25, 9-18.

55

56

SPECIAL POPULATIONS
Chapter 3: High-Risk and High-Cost Offenders in Proposition 36
Angela Hawken, Ph.D.
The arrest and court costs that accumulate when offenders commit new crimes are a
significant driver of follow-up costs under Prop 36. The typical (median) Prop 36
offender contributes little to arrest and court costs, while a small number of offenders
contribute disproportionately to these costs. 25% of Prop 36 offenders account for 80%
of follow-up crime costs. Only 14% of those high-cost offenders who entered treatment
had a successful treatment completion.
Prop 36 participant demographics were not strong predictors of follow-up recidivism.
High crime cost offenders had the same race/ethnic profile as lower cost offenders, were
more likely to be male, and were, on average, about three years younger.
There were no meaningful differences in the drug treatment histories of high cost
offenders compared with the general population of Prop 36 offenders, and this group was
as likely to enter treatment under Prop 36. Treatment completion rates were lower for
this group than the general population (their follow-up crime rates were higher, and as a
result they had higher rates of incarceration).
A strong predictor of follow-up recidivism is the number of convictions in the 30 months
preceding an offender’s entry into Prop 36. The per-day on street arrest and conviction
costs are twenty-six times higher for those with five or more prior convictions than for
those who enter with no prior convictions. Crime costs increase monotonically with the
number of prior convictions (i.e., the average crime cost increases as the number of prior
convictions increase). The average daily crime cost of Prop 36 offenders with no prior
convictions was $28 per day. The average daily crime cost of Prop 36 offenders with five
or more convictions was $723 per day.
Prop 36 offenders with long criminal histories are readily identifiable and a number of
risk assessment tools are available. UCLA recommends identifying high-risk offenders
early, and managing them differently. Possible responses might include: making highrisk offenders ineligible for Prop 36, particularly if their convictions include non-drug
crimes; incapacitating these offenders during their participation in Prop 36 by requiring
residential treatment; or intensively supervising these offenders while under community
supervision. Responding appropriately to the supervision and treatment needs of highrisk participants will be a challenge given the limited funding available for Prop 36.
Arrest and court costs accumulate, public safety is undermined, and the reputation of Prop 36
is hurt when offenders commit new crimes. Here UCLA considers the distribution of the
costs of arrests and convictions, describe Prop 36 participants who are classified as high-cost
offenders, and propose policies to manage offenders deemed at high risk of costly
recidivism.

57

Distribution of Crime Costs
There is a great deal of variation in the individual contributions towards total crime costs
among Prop 36 participants. The typical (median) Prop 36 offender contributes very little
to arrest and court costs, while a small number of offenders contribute disproportionately
to these costs.
Figure 3.1 illustrates the distribution of arrest and court costs and shows the inequality of
contribution to total crime cost across offenders. Offenders are rank-ordered based on their
contribution to overall costs (ordered from low to high). The x-axis shows the percentage of
offenders and the y-axis shows the percentage of arrest and court costs attributed to those
offenders.1 The diagonal line indicates the scenario in which all offenders contribute equally
to overall arrest and conviction costs. The curved line shows the actual contributions. The
figure indicates a highly unequal distribution of arrest and convictions costs: eighty percent
are contributed by twenty-five percent of offenders.

Cumulative Proportion of Total Crime Costs

Figure 3.1
Inequality of Crime Cost Contributions

Cumulative Proportion of Offenders
Note: Data for arrests and convictions are from the Department of Justice Automated Criminal
History System. Crime costs are adjusted (see Hawken et al, 2007). Offenders are rank-ordered
based on their contribution to overall costs (ordered from low to high). The x-axis shows the
percentage of offenders and the y-axis shows the percentage of arrest and court costs contributed by
those offenders.

The majority of Prop 36 follow-up recidivism involves drug crimes. 54% of offenders that
were eligible for Prop 36 during its first year, have a new drug arrest with in the 42 months
following their entry into Prop 36, 21% have a new arrest for property crimes, and 7% have a
1

This is similar to a Lorenz Curve, which shows income inequality across households. The graphic here shows
the inequality in contribution to overall crime costs.

58

new arrest for violent crimes. But the crimes committed by high cost offenders include
many non-drug crimes. Of particular concern for public safety is the large number of violent
crimes (including assault, rape, homicide, and in some instances multiple homicides)
committed by this group.2
The obvious next step is to identify which offenders are on the right-hand side of the
distribution (contribute a large percentage of costs of new crimes) and which are on the left
(little-to-no costs contributed). In the analysis that follows the focus is on offenders in the
top end of the crime cost distribution. This group accounts for 80% of overall crime costs.
Who are these high cost offenders?

Identifying high-cost offenders
Here high-cost offenders are described. Participant demographics, prior treatment histories,
treatment under Prop 36, and prior criminal histories are summarized.
Demographics
An analysis of characteristics related to follow-up crime costs shows that demographic
characteristics are poor predictors of court and arrest costs in the follow-up period.
Demographic characteristics of offenders in the upper and lower quartiles of the cost
distribution are reported in Tables 3.1- 3.3.
Race/ethnicity
Table 3.1 shows the race/ethnic composition of low-cost and high-cost Prop 36 participants.
There is no meaningful difference in the distribution of race/ethnicity between highest- and
lowest-quartile offenders.

Table 3.1: Comparison of Race/ethnicity of Prop 36 Offenders in the Lowest and
Highest Quartiles of Contribution to Arrest and Conviction Costs
Lowest Quartile
Highest Quartile
Black
16.1%
16.0%
Hispanic
32.3%
33.3%
White
46.9%
47.1%
Other
4.7%
3.6%
Note: Data for arrests and convictions are from the Department of Justice Automated Criminal History
System. Crime costs are adjusted (see Hawken et al., 2007).

Sex
Table 3.2 distinguishes low-cost and high-cost offenders by sex. Males were overrepresented in the group of high-cost offenders, with 5% more males in the high quartile than
in the lower quartile. But sex alone was not a strong predictor of follow-up recidivism.

2

The criminal justice literature shows that for most types of crimes, only a small percentage of actual crimes
committed result in an arrest. The ratio of actual crimes committed for each arrest differ by crime type. This
analysis considers only reported arrests and convictions, and does not capture crime costs attributable to an
offender if the crime did not lead to an arrest.

59

Table 3.2: Comparison of sex of Prop 36 Offenders in the Lowest and Highest
Quartiles of Contribution to Arrest and Conviction Costs
Lowest Quartile
Highest Quartile
Male
73.5%
78.5%
Female
26.5%
21.5%
Note: Data for arrests and convictions are from the Department of Justice Automated Criminal History
System. Crime costs are adjusted (see Hawken et al, 2007).

Age
The criminal justice literature shows a strong relationship between age and criminal activity;
with younger adults having higher rates of criminality. Table 3.3 shows the average age of
the low-cost and high-cost offender groups. There was a slight difference in age between
high-cost offenders, and the general population of Prop 36 offenders. High cost offenders
were 3.5 years younger on average than those with lower follow-up crime costs.

Table 3.3: Comparison of ages of Prop 36 Offenders in the Lowest and Highest
Quartiles of Contribution to Arrest and Conviction Costs
Lowest Quartile
Highest Quartile
Age
35.5
31.9
Note: Data for arrests and convictions are from the Department of Justice Automated Criminal History
System. Crime costs are adjusted (see Hawken et al, 2007).

High Cost Offenders and Treatment
Here UCLA compares the treatment experience of high-cost Prop 36 offenders and the full
population of Prop 36 participants. The goal was to determine whether prior treatment
exposure or treatment participation under Prop 36 mitigated high-cost criminality for this
group.
Prior treatment
A recent history of treatment did not mitigate differences in follow-up high-cost recidivism.
There was no meaningful difference in the recent treatment history of high cost offenders
and the general Prop 36 population. Just over a quarter of the offenders in both groups had
been admitted to treatment during the 30 months preceding their entry into Prop 36.

Table 3.4: Treatment status of High-Cost and General Prop 36 Offenders
Treatment in Prior 30 months
Treated in Past 30 months
High-Cost Offenders
All Prop 36
Drug Treatment
27.1%
26.3%
Untreated
72.9%
73.7%
Note: Data are from CADDS.

60

Prop 36 Treatment Entry
Offenders eligible for Prop 36 who go on to become high-cost offenders were as likely to
have been treated under Prop 36 as the general population of Prop 36 participants.
Prop 36 Treatment Completion
Completion rates for high-cost offenders were low. Only 14% of those high-cost offenders
who entered treatment had a successful treatment completion discharge.3 There are two
likely explanations for the low treatment completion rates found for the high-cost group.
High-cost offenders are more likely to drop out of treatment and high-cost offenders are
more likely to be re-incarcerated, thereby disrupting treatment.
High Cost Offenders and Prior Convictions
The best predictor of follow-up recidivism was the offenders’ number of convictions in the
30 months preceding their entry into Prop 36. Figure 3.2 shows the probability of being in
the upper two quartiles of crime costs, given the number of prior convictions accumulated.
The likelihood that a Prop 36 offender would become a high-cost offender increases
monotonically with the number of prior convictions. 62% of offenders with five or more
prior convictions went on to become high-cost crime offenders (upper quartile), compared
with 18%, of those entering Prop 36 with no prior convictions.
A similar pattern holds for number of prior convictions and follow-up crime costs. A solid
predictor of offender arrest and conviction costs in the follow-up period was the number of
convictions the offender incurred in the 30-month period prior to the SACPA-eligible
conviction. Arrest and conviction costs increase monotonically as the number of prior
convictions increases. There is a marked increase in costs between the group with four or
fewer prior convictions and those with five or more. Figure 3.3 illustrates this difference.
Those with five or more convictions in the 30-month period before their Prop 36-eligible
convictions, constituting 1.6% (N = 1,010) of the Prop 36 group, had post-conviction crime
costs in the 30-month follow-up period ten times higher than the typical Prop 36 offender
($21,175 versus $2,254).
The crime cost differential is even more dramatic when days-at-risk are taken into account,
as most high-cost offenders are re-arrested and re-incarcerated. A comparison of costs per
offender per “day on the street” illustrates the strong relationship between the number of
prior convictions (zero to four, and five or more) in the 30 months prior to the Prop 36eligible conviction and follow-up costs (see Figure 3.4). The average daily crime cost of
Prop 36 offenders with no prior convictions was $28 per day. The average daily crime cost
of Prop 36 offenders with five or more prior convictions was $723 per day. The crime cost
differences reported here capture only arrest and court costs. The crime differential for the
full social cost of crimes committed (these would include victims costs and non-pecuniary
costs to society) would be substantially greater as the crimes committed by high-cost
offenders include more property and violent crimes.

3

Not all high-cost offenders who were referred to treatment under Prop 36 entered treatment. The treatment
completion rate over all high-cost offenders referred to treatment (including those who entered treatment and
those who were no-shows) was less than 8%.

61

Percentage of offenders

Figure 3.2
Follow-up Crime Category and Prior Convictions
100%
80%
60%
40%
20%
0%
0

1

2

3

4

5+

Number of Prior Convictions in Past 30-Months
Below Median

3rd Quartile

4th Quartile

Note: Data are from the California Department of Justice. Crime costs are
adjusted (see Hawken et al., 2007)

Figure 3.3
Relative Costs for Offenders with 5+ Prior Convictions
(1.6%, N = 1,010)
25000

Dollars per offender

$21,175
20000
15000
10000
5000
$2,254
0
Typical Offender

5+ Prior Convictions

Note: Data for arrests and convictions are from the Department of Justice
Automated Criminal History System. For details on costs assigned to each
arrest and conviction see Hawken et al. 2007. Offenders with five or more prior
convictions constitute 1.6 percent of the Prop 36 sample (N = 1,010).

62

Figure 3.4
Arrest and Conviction Costs Per Day on Street by Number of
Prior Convictions

$723

$800

Dollars per day on street

$700
$600
$500
$400
$300

$166

$200
$100

$28

$51

0

1

$218

$103

$0
2

3

4

5+

Number of prior convictions
Note: Data for arrests and convictions are from the Department of Justice Automated
Criminal History System. For details on costs assigned to arrests and convictions
(see Hawken et al., 2007).

Stakeholder Responses on Practices to Manage High-Risk Offenders
Prop 36 participants who enter with a large number of prior convictions have a high
probability of high-cost recidivism. Given this risk, are counties managing these offenders
differently?
The UCLA Lead Agency Surveys solicited responses from counties on techniques currently
used to manage offenders who enter Prop 36 with a large number of known priors. Figure
3.5 shows the practices reported. 34% of counties have no strategies in place to supervise
high-risk offenders more closely, or to provide them with more-intensive treatment services.
59% of counties reported that offenders entering Prop 36 with a large number of prior
convictions were subject to increased monitoring and supervision. 37% of counties reported
that Prop 36 clients entering with a large number of prior convictions were more likely to be
assessed for residential treatment. 32% of counties used a combination of more-intensive
monitoring for high-risk offenders and prioritizing these offenders for residential treatment.
A number of counties noted that while high-risk offenders were prioritized for residential
care, placement was conditional on the availability of residential care; as a result, not all
high-risk offenders could be accommodated. Only two counties reported additional

63

strategies to manage high-risk offenders. These strategies included intensive case
management4, and requiring more court appearances and contact with the court.

Figure 3.5
Lead Agency Report of Practices used to Manage High-Risk
Offenders
70%
59%

60%
50%
40%

37%

34%

30%
20%
10%
0%
No strategies used

Increased supervision Residential placement

Note: Data are from the 2007 UCLA Stakeholder Survey data collected from the
Lead Agencies. See Appendix A for a description of the survey.

Risk assessment
Individuals entering Prop 36 with five or more prior convictions were shown to be high-risk
offenders, but many offenders with fewer priors also go on to high-cost recidivism. A key to
effectively managing high-risk offenders is identifying who they are and sharing this
information with those individuals involved in managing the offender (in particular, the
treatment provider and probation officer). Identifying high-risk offenders would require
expanding the use of risk assessment tools, whereby Prop 36 offenders are assessed on key
factors known to be associated with high-cost recidivism.5
There have typically been three approaches to risk assessment for criminal offenders
(corresponding to three generations of risk assessment literature): clinical opinion, actuarial
prediction, and structured-professional judgment.6

4

There are many competing definitions of intensive case management. Respondents did not specify their
definition of intensive case management.
5
The state of Kansas has a treatment diversion program similar to Prop 36. The state oversaw the creation of a
team of community supervision specialists who conduct risk assessment on every program participant. Kansas
uses the LSI-R throughout their system to determine the appropriate supervision level.
6
For a review of the risk assessment literature, classified into three distinct generations see Bonta 1996.

64

The first generation of risk assessment primarily involved unstructured judgments made by
clinical practitioners. This approach has been largely discredited due to the subjective nature
of the assessments and their poor predictive power (Hannah-Moffit, 2005).7
The second generation of risk assessment tools were developed in the 1970s and relied on
evidence-based actuarial prediction. These prediction models relied exclusively on static
historic risk factors such as age and prior criminal history. These models had much greater
predictive power than earlier methods, but were later criticized for their lack of flexibility,
their excessive reliance on static offense-based criteria, and their inability to help design
targeted interventions (Andrews & Bonta, 1998, Hannah-Moffit, 2005).
The third generation risk assessment tools, known as Structured Professional Judgment,
combine elements of the first and second generation approaches. This approach requires
practitioners to add dynamic factors (“criminogenic needs” such as anti-social personality,
poor self-control, family dysfunction, and lack of unemployment or vocational skills) to the
assessment, including variables such as employment, family relationships, personality traits,
and attitudes (Andrews & Bonta, 1998). This approach allows practitioners to take casespecific individual details into account. Third generation assessment tools are claimed to
objectively and systematically measure both static and dynamic risk (Hannah-Moffit, 2005).
The Level of Service Inventory-Revised (LSI-R) is the most commonly used assessment
instrument. It is a third generation assessment tool that uses a structured professional
judgment assessment approach. The assessment covers 54 items on a wide array of risk
factors:
•

Antisocial attitudes

•

Antisocial thoughts, cognitions and ways of thinking

•

Antisocial personality

•

Antisocial history

•

Employment

•

Family

•

Leisure and recreational activities

•

Substance abuse problems

•

Antisocial peers or criminal associates

Other assessment tools are currently under review in California, including Correctional
Offender Management Profiling for Alternative Sanctions (COMPAS) and Risk and Needs
Triage (RANT developed by Douglas Marlowe). COMPAS is a product of Northpointe
Institute for Public Management. The tool relies on official records and self-reported data,
and captures a broad array of risk factors, selected for their ability to predict recidivism and
7

On average, these assessment tools performed no better than chance at predicting outcomes.

65

compliance with terms of community supervision. COMPAS has been used in many
contexts, but most relevant to Prop 36, it has been applied to probation supervision.
The RANT model was designed specifically for drug-involved offenders. The tool identifies
risk levels and matches offenders to levels of supervision and treatment that are proven to be
cost-effective given the specific characteristics of the offender.
The California Department of Alcohol and Drug Programs should encourage counties to use
assessment tools and should implement pilot studies to determine the relative predictive
powers of these tools when applied to the Prop 36 population.

Recommendations for Managing High-Cost Offenders
Counties should work with the criminal justice and substance abuse treatment communities
to identify strategies to manage those offenders deemed to be at high risk for costly reoffending. Options in this regard include: providing more intensive probation supervision,
requiring residential care for these offenders, or revising Prop 36 eligibility criteria.
Exclude high-risk offenders from participating in Prop 36
•

Advantage: High cost offenders who commit violent crimes are a threat to
public safety and hurt the image of the Prop 36 program. Sentencing these
offenders outside of Prop 36 may be more appropriate for these types of
offenders. These offenders may be better served under the close supervision
of a drug court judge.

•

Disadvantage: Excluding high-risk offenders from Prop 36 eligibility would
require a revision to the Prop 36 law.

•

Disadvantage: While a high proportion of offenders with many priors go on
to high-cost recidivism, not all offenders with many priors are a risk to public
safety. Exclusion criteria become a concern. For example, it may be
appropriate to distinguish eligibility based on the nature of prior convictions.
Offenders who enter with prior drug offenses only may benefit from Prop 36,
even if they have accumulated a number of priors. Those who enter with
prior convictions that include violent crimes pose a greater risk to public
safety. Determining the relevant criteria for exclusion from Prop 36 may be
controversial and would require input from many stakeholder groups.

Residential treatment.
•

Advantage: High-risk offenders are incapacitated, reducing the risk they pose
to public safety. High-risk offenders placed into residential care had lower
follow-up crime costs than those allocated to outpatient or methadone
maintenance.

•

Disadvantage: Treatment completion rates for high-cost Prop 36 clients were
extremely low. Community-based residential treatment may not be sufficient
to incapacitate offenders who pose a great risk of high cost recidivism. A
treatment approach alone cannot solve the problem of high-cost offenders.

66

Even those who completed residential treatment did less well than lower-risk
offenders.
•

Disadvantage: Prop 36 residential beds are extremely scarce. High-cost
offenders would displace other Prop 36 offenders who may be more
amenable to treatment and who would benefit more from the residential
services provided.

Intensive community supervision
•

Advantage: The literature shows improvements in offender outcomes for
intensive supervision (including more-frequent reporting and drug testing)
coupled with drug treatment.

•

Disadvantage: Intensive community supervision would require reducing
probation and parole caseloads, and increasing reporting and drug testing
requirements. Given the uncertainty associated with Prop 36 funding, and
the recent cuts to Prop 36 funding, counties may be wary of longer term
commitments (such as hiring and training new probation officers), and would
be hard-pressed to find the resources needed to cover the costs of intensive
supervision.

Any revisions to Prop 36 would have implications for the cost of operating the program.
Amendments to the current law would need to be carefully evaluated based on how these
changes are likely to affect outcomes, and the costs of operation.

References
Andrews, D., & Bonta, J. (1998). Psychology of Criminal Conduct. Cincinnati, OH:
Andersen Publishing.
Hannah-Moffat, K. (2005). Criminogenic needs and the transformative risk subject:
Hybridizations of risk/need in penalty. Punishment and Society, 7.
Hawken, A., Longshore, D., Urada, D. & Anglin, M.D. (2007). SACPA Benefit-Cost
Analysis. In Evaluation of the Substance Abuse and Crime Prevention Act: Final
Report. Los Angeles: University of California, Los Angeles.

67

68

Chapter 4: The Homeless Mentally Ill in Proposition 36
Bradley T. Conner, Ph.D. and Christine Grella, Ph.D.
Research shows that somewhere between 55% and 69% of individuals diagnosed with an
alcohol or drug use disorder have also been diagnosed with a co-occurring mental health
disorder. This population is also more likely to be homeless. The goal of this chapter is to
determine the prevalence and differential outcomes of Prop 36 clients who are homeless
and have co-occurring drug use and mental health disorders.
Drug treatment providers reported that, on average, 20.6% of their clients were homeless
and had a co-occurring mental disorder at treatment entry. However, 37.5% of lead
agencies reported that they did not conduct a mental health screening during the assessment
process, 28.6% of Prop 36 providers reported employing mental health professionals, and
54.2% of lead agencies reported offering housing services to the homeless Prop 36 clients.
California Department of Mental Health (DMH) administrative data indicate that Prop 36
clients that also received mental health services in the 12 months following their Prop 36
conviction spent significantly fewer days in drug treatment and that those that were both
homeless and receiving mental health services spent the fewest days in treatment when
compared to those not homeless and not identified in DMH administrative data. Homeless
offenders eligible for Prop 36 who were also receiving mental health services were more
likely to get arrested for drug, property and violent crimes than the comparison groups in
the 30 months following the conviction that made them eligible for Prop 36 participation,
indicating that this is a very difficult population to treat effectively.
Integrated Dual Diagnosis Treatment (IDDT) is an evidence-based treatment of cooccurring disorders that the Federal Substance Abuse and Mental Health Services
Administration recommends as the preferred treatment for co-occurring disorders. Finding
ways to build IDDT into the current treatment regimen may improve outcomes associated
with the treatment of Prop 36 clients who are homeless and have co-occurring disorders.
California has two separate departments with two separate means for funding drug and
mental health treatment. ADP’s funding includes Prop 36, while DMH’s includes Prop 63.
Both Prop 36 and Prop 63 include a focus on serving homeless clients that may have mental
health disorders including those for alcohol and other drugs disorders. This separation
seems to be an impediment to integrating care for the homeless mentally ill drug users in
Prop 36. UCLA is recommending integration of these two sources of money to allow for
the creation of “Whatever It Takes” approaches to treating these difficult clients. This
could be accomplished by awarding Prop 36 contracts and Prop 63 grants to IDDT
facilities.
Research shows that 55% to 69% of individuals with a substance (alcohol or drug) use
disorder have a co-occurring mental heath disorder (see Watkins et al., 2004 for review).
Research has also shown that as many as 60% of those individuals who have been diagnosed
with a mental disorder also have co-occurring substance use disorder (Mueser et al., 2003;

69

Reiger et al., 1990). Co-occurring disorders may include any combination of two or more
substance use disorders (e.g., alcohol abuse or dependence, cocaine abuse or dependence,
polysubstance abuse or dependence) and mental disorders (i.e., major depression,
schizophrenia, or posttraumatic stress disorder) as defined in the Diagnostic and Statistical
Manual of Mental Disorders–IV–Text Revision (DSM-IV-TR; American Psychiatric
Association, 2000). There are no specific combinations of substance use and mental
disorders that are uniquely defined as co-occurring disorders, however, some combinations
are more prevalent than others, for example alcohol use and mood disorders (Conway et al.,
2006; Kessler, 2004).
Several epidemiological studies have reported the prevalence of co-occurring mental and
substance use disorders (for example see the National Comorbidity Study and its replication
and the National Epidemiologic Survey on Alcohol and Related Conditions). Empirical
research indicates that individuals with co-occurring disorders are more likely to be arrested,
incarcerated, and spend more time incarcerated than those without mental disorders (Drake et
al., 2001; Monahan et al., 2005). Additionally, this group of individuals has more trouble
getting and keeping employment or other forms of financial support, reliable transportation,
or appropriate medical and mental health care (Brunette & Mueser, 2006). This population
is also more likely to be homeless, or have varying patterns of residential instability,
including precarious and unstable housing (e.g., “doubling up” with others, living in hotels
and motels), intermittent homelessness (i.e., periodic shelter use), and chronic homelessness
(Osher & Dixon, 1996). Research indicates that those that are homeless are also more likely
to be imprisoned (Kushel et al., 2005). Additional research has documented that homeless
mentally ill drug users typically do less well than their counterparts without mental illness or
those with stable and suitable housing on a number of psychosocial outcome measures
(Gonzalez & Rosenheck, 2002; Zuvekas & Hill, 2000). All of these factors combine to make
it very difficult to track, study, and treat this special population.
Homeless individuals (or those who are at risk of homelessness) with co-occurring disorders
are likely to have frequent contact with the criminal justice system because they typically
cycle through acute care facilities in the community, such as hospital emergency rooms or
crisis units, and, lacking stabilization or long-term support, eventually wind up in jail or
prison (Peters et al., 2004). In one study over 75% of homeless inmates with a severe mental
disorder had a co-occurring substance use disorder; these inmates were more likely to be
homeless and to be charged with violent crimes than other inmates (McNiel et al., 2005). A
recent review of research comparing offenders with mental disorders only and those with cooccurring disorders found that those with co-occurring disorders were more likely to be
serving sentences related to their substance use, to be homeless, to violate probation after
release, and to recidivate to correctional custody (Hartwell, 2004).
Several major societal changes in the past 50 years have combined to increase the risk of
homelessness among individuals with co-occurring disorders. These include decreasing
housing capacity, particularly in urban areas, the fragmentation of public health systems,
including both mental disorders and substance use disorders treatment, and the proliferation
and easy availability of alcohol and illicit drugs that have permeated society (Caton, 1990;
Rossi, 1989). Hence, interventions aimed at addressing this population need to consider their
multiple and intersecting problems, which require coordination of services across multiple

70

service sectors that are frequently lacking in capacity or resources (Lamberti et al., 2001).
Research has shown that homeless individuals with co-occurring disorders often find it
difficult to access treatment services, despite their high levels of need (Wenzel et al., 2001).
The barriers to providing coordinated treatment across service systems, as well as integrated
treatment, have been well documented (McGovern et al., 2006).
The goal of this chapter is to illustrate the prevalence among Prop 36 clients of the
homelessness and co-occurring mental disorders in the population of offenders in the state of
California who decided to participate in Prop 36. As Prop 36 does not extend to convictions
for alcohol use disorders, this chapter will focus on co-occurring drug use and mental
disorders. In addition to providing prevalence data, differential performance and outcomes
between homeless with co-occurring disorders and those with stable housing and no mental
illness will be discussed. Finally, recommendations for how to improve the Prop 36 program
for this population throughout the supervision and treatment process will be discussed.

Homelessness and Co-Occurring Disorders in Prop 36
Mental Disorder Assessment in Prop 36
Data presented in this chapter were derived from a number of sources including the UCLA
Stakeholder and Program Surveys (completed by lead agencies, courts, probation, parole,
and treatment providers across the state) focus groups made up of various stakeholders
conducted across the state (Appendices A through C provide detailed information on the
methods used to collect, maintain, and analyze this data), and administrative databases from
the California Department of Justice (DOJ), the California Department of Mental Health
(DMH), and the California Alcohol and Drug Data System (CADDS).
As part of the Prop 36 “pipeline” each offender that elects to participate in Prop 36 must
undergo an assessment. The assessment is to be used to inform treatment recommendations.
On the UCLA Stakeholder Survey lead agencies indicated if a mental disorder screening1
was either 1) routinely conducted as part of this assessment or 2) if a mental disorder
screening was conducted in response to elevations on items of a typical assessment (i.e., the
Addiction Severity Index or the American Society of Addiction Medicine Patient Placement
Criteria) that indicates that the individual may be suffering from a mental disorder. Of the
48 lead agencies that responded to the survey, 31.25% (15) reported routinely conducting a
mental disorder screening in addition to the typical assessment instrument and 37.5% (18)
reported conducting a mental disorder screening in response to elevations on the typical
assessment that indicated the presence of a mental disorder, meaning that 37.5% (18) of the
lead agencies reported that they did not conduct a mental disorder screening as part of the
assessment (3 agencies reported conducting both types of mental disorder assessment).
Table 4.1 details the methods used to address the presence of a mental disorder in the
participant being assessed by those lead agencies that reported conducting any type of mental
disorder screening.

1

An assessment is typically a detailed interview for the purpose of diagnosis, classification, or service planning
whereas a screening is a brief inquiry as to whether a problem exists. Typically a positive result on a screening
should lead to a detailed assessment.

71

Table 4.1: Methods for Addressing Mental Disorders Detected at Assessment
Method
Assignment to a treatment program specializing in the treatment
of co-occurring disorders
Referral for mental health services at a provider other than the
drug use disorders treatment provider
Referral to a licensed mental health professional

N

%

17

56.7%

26

86.7%

25

83.3%

Note: Percentages are based on the 30 lead agencies (62.5%) that reported conducting a mental disorders
assessment.

Mental Disorder Frequencies in Prop 36
Analysis of the fiscal year (FY) 05/06 CADDS data provides a limited indication of the
prevalence of mental disorders in the Prop 36 population who entered drug treatment.
Individuals completing a CADDS assessment at drug treatment entry are asked to identify
disabilities. According to CADDS data, 4.14% (1,661) of Prop 36 clients reported having a
mental disability at drug use disorders treatment entry. There is an optional item on the
CADDS assessment that asks “Has this participant ever been diagnosed as also having
chronic mental illness?” however, the response frequency is very low (around 20% of Prop
36 client CADDS forms had any response on this item), indicating that about 2% (819) of
Prop 36 clients had ever been diagnosed with a chronic mental illness. These items are
limited because they are based on either client self-report or an estimation by the person
completing the CADDS form, who is likely not a mental health professional and thus does
not have the training to make a diagnosis. It should be noted that ADP is switching to a new
data system –CalOMS –that should provide somewhat better data on AOD treatment clients
with co-occurring needs.
UCLA was able to access data from DMH for those clients that received mental health
services in addition to treatment for their drug use disorders. It should be noted that
California’s DMH is charged with treating only those individuals with severe and persistent
mental illness, so these numbers only reflect the most severe individuals, it is likely that
there are many more individuals receiving drug treatment through Prop 36 who have a cooccurring mental health disorder that do not meet this criteria. Approximately 24% of all
Prop 36 clients were present in the DMH data reporting system, however, on average 8%
received mental health services concurrently with the drug treatment for their Prop 36
eligible conviction. This gives an approximation of the percentage of Prop 36 clients who
were formally diagnosed with mental disorders and were referred and received mental health
treatment through the public mental health system, both on the whole and during their
participation in Prop 36. Additionally, individuals who are able to pay for mental health
services with private insurance or out-of-pocket are likely not present in the DMH
administrative data. Finally, the inherent difficulties associated with matching administrative
data may also have limited UCLA’s ability to accurately identify all clients receiving mental
health and drug treatment concurrently while completing their participation in Prop 36.
Though multiple funding sources are available, the primary mechanism used to fund the
public mental health system in California is Medi-Cal. California residents may be eligible
for Medi-Cal if they receive assistance from the Supplemental Security Income/State

72

Supplemental Program (SSI/SSP), the California Work Opportunity and Responsibility to
Kids (CalWORKs, previously called Aid to Families with Dependent Children or AFDC), or
Refugee Assistance, if they participate in the Foster Care or Adoption Assistance Program, if
they are 65 or older, blind, disabled, under the age of 21 years, pregnant, diagnosed with
breast or cervical cancer, in a skilled nursing or intermediate care facility, or if they have
refugee status during a limited period of eligibility. California residents are also eligible to
receive Medi-Cal if they are a parent or caretaker relative of a child under 21 and the child’s
parent is deceased or doesn’t live with the child, or the child’s parent is incapacitated, or the
child’s parent, who is the primary wage earner, is unemployed or underemployed, meaning
that the family is below the Federal Poverty Line (a comprehensive list of eligibility criteria
are presented in the Medi-Cal eligibility manual).
Using the DMH administrative data, UCLA can detail the distribution of the most common
diagnoses in the Prop 36 population. The most common diagnosis, at 8.1% of the 24%
present in both the DOJ data and the DMH data was Depressive Disorder Not Otherwise
Specified (NOS). See Tables 4.2-4.4 for more information regarding the distribution of
diagnoses. The majority of diagnoses were from the family of Mood Disorders (37.2%),
Substance Use Disorders (22.8%), or the Schizophrenias (16.5%).

Table 4.2: Distribution of Mental Disorders in Prop 36
Disorder

Frequency

Percent

Depressive Disorder NOS

6,125

8.1%

Psychotic Disorder NOS

5,947

7.8%

Mood Disorder NOS

5,598

7.4%

Diagnosis Deferred

3,551

4.7%

Polysubstance Dependence

2,909

3.8%

Schizoaffective Disorder

2,385

3.0%

Schizophrenia, Paranoid Type

2,207

2.9%

Adjustment Disorder Unspecified

2,025

2.7%

Bipolar Disorder NOS

1,940

2.6%

Amphetamine Dependence

1,816

2.4%

Total

34,503

45 %

Note: Only the top 10 diagnoses are given in the table as the total list is too long to reproduce here. There are a
total of 10,509 individuals represented in the table above. The disorders are based on DSM-IV-TR diagnoses.

It should be noted that these data were obtained from administrative databases and, as such,
are limited as they may contain errors potentially affecting reliability and validity as a result
of problems during data collection, entry, and internal maintenance processes. Additionally,
data are presented for unique individuals (each person) but each person can have more than
one mental health treatment episode and multiple diagnoses. Individual data by Prop 36 year
is presented later in this chapter.

73

Table 4.3: Distribution of Diagnoses by Subcategory
Subordinate Diagnostic Category

Frequency

Percent

Drug Use Disorders

15,057

19.9%

Major Depression

8,032

10.6%

Mood Disorder

6,932

9.1%

Bipolar Disorder

6,709

8.9%

Depressive Disorders

6,125

8.1%

Psychotic Disorder NOS

6,099

8.0%

Adjustment Disorders

5,810

7.7%

Diagnosis Deferred

3,551

4.7%

Schizoaffective Disorders

2,410

3.2%

Anxiety Disorders

2,329

3.1%

Schizophrenia Paranoid Type

2,291

3.0%

Alcohol Use Disorders

2,207

2.9%

Stress Disorders

1,218

1.6%

Schizophrenia Undifferentiated Type

1,192

1.6%

877

1.2%

70,839

93.4

Conduct Disorders

Total

Note: The total, 70,839, represents the number of unique diagnoses present, not individuals. There are 21,818
individuals represented in the table. The remaining 6.57% of diagnoses not present in the table represented less
than 1% each of the data and were too numerous to list here. The subordinate diagnostic categories are based
on the major divisions of the 16 major diagnostic classes of the DSM-IV-TR.

As data from individual diagnoses are not easily summarized, primarily due to the large
number of different diagnoses, from hereon only Diagnostic Class Data will be discussed.
Diagnostic classes are the broadest grouping of disorders that typically represent a type of
disorder, as in the mood disorders, or a developmental period, as in disorders usually
diagnosed in childhood. This aggregated data will ease communication of the findings.
As all previous UCLA evaluations of Prop 36 have been by fiscal year and as UCLA was
interested in differences in outcomes for individuals who received mental health services
while also receiving drug treatment through Prop 36, DMH data are also presented in this
yearly form. The diagnostic data for Prop 36 eligible offenders receiving mental health
services by year are presented in Tables 4.5-4.6.

74

Table 4.4: Distribution of Diagnoses by Diagnostic Class
Superordinate Diagnostic Category

Frequency

Percent

Mood Disorders

28,202

37.2%

Substance Use Disorders

17,267

22.8%

Schizophrenia/Psychotic Disorders

12,472

16.5%

Adjustment Disorders

6,320

8.3%

Diagnosis Deferred

3,551

4.7%

Anxiety Disorders

3,052

4.0%

Disorders Usually Diagnosed in Childhood

1,517

2.0%

V Codes

1,179

1.6%

Personality Disorders

469

0.6%

Impulse Control Disorders

424

0.6%

Disorders due to a General Medical Condition

409

0.5%

Mental Disorder NOS

284

0.4%

Cognitive Disorder NOS

238

0.3%

Malingering

221

0.3%

Sleep Disorders

45

0.06%

Somatoform Disorders

31

0.04%

Eating Disorders

30

0.04%

Dissociative Disorders

28

0.04%

Sexual Disorders

25

0.03%

Problems with Physical Abuse

24

0.03%

Factitious Disorders

11

0.01%

Medication Induced Disorders

10

0.01%

Noncompliance with Treatment

8

0.01%

75817

100.00

Total

Note: The total, 75,871, represents the number of unique diagnoses present, not individuals. There are a total of
23,352 individuals represented in the above table. The superordinate diagnostic categories are based on the 16
major diagnostic classes of the DSM-IV-TR.

75

Table 4.5: Distribution of Diagnostic Classes in Year 1 of Prop 36 (n = 3403)
Superordinate Diagnostic Category
Frequency
Percent
Mood Disorders
1,337
39.3%
Schizophrenia/Psychotic Disorders
713
21.0%
Substance Use Disorders
698
20.5%
Adjustment Disorders
242
7.1%
Anxiety Disorders
151
4.4%
Diagnosis Deferred
92
2.7%
V Codes
60
1.8%
Disorders Usually Diagnosed in Childhood
36
1.1%
Disorders due to a General Medical Condition
16
0.5%
Impulse Control Disorders
15
0.4%
Personality Disorders
14
0.4%
Mental Disorder NOS
12
0.4%
Malingering
6
0.2%
Cognitive Disorder NOS
5
0.1%
Dissociative Disorders
2
0.06%
0.01%
Eating Disorders
1
0.01%
Medication Induced Disorders
1
0.01%
Sexual Disorders
1
Sleep Disorders
1
0.01%
Table 4.6: Distribution of Diagnostic Classes in Year 2 of Prop 36 (n = 3869)
Superordinate Diagnostic Category
Frequency
Percent
Mood Disorders
1467
37.9%
Substance Use Disorders
796
20.6%
Schizophrenia/Psychotic Disorders
768
19.9%
Adjustment Disorders
289
7.5%
Anxiety Disorders
198
5.1%
Diagnosis Deferred
171
4.4%
V Codes
65
1.7%
Disorders Usually Diagnosed in Childhood
41
1.1%
Mental Disorder NOS
18
0.5%
Disorders due to a General Medical Condition
15
0.4%
Impulse Control Disorders
14
0.4%
Personality Disorders
13
0.3%
Cognitive Disorder NOS
4
0.1%
Dissociative Disorders
3
0.08%
Somatoform Disorders
2
0.07%
Abuse
1
0.01%
0.01%
Eating Disorders
1
Factitious Disorders
1
0.01%
Malingering
1
0.01%
Sleep Disorders
1
0.01%

76

In year 1, FY 01/02, 3,403 of the 40,368 Prop 36 eligible offenders received mental health
services during the 12 months following their eligible conviction, representing
approximately 8.4% of that year’s Prop 36 eligible population. This number increased to
3,869 of 41,578 in year 2, FY 02/03, representing 8.5% of that year’s Prop 36 population.

Homelessness in the Prop 36 Population
While the data on the mental disorders in the Prop 36 population is interesting, the focus of
this chapter is on the special population of Prop 36 clients that are both homeless and have a
co-occurring mental disorder. As such, demographic data from the 2007 UCLA Program
Survey and CADDS on homelessness will be presented first, followed by data on the overlap
of homelessness and mental disorders in the Prop 36 population identified and matched
across three administrative databases: DOJ, DMH, and CADDS.
The UCLA Program survey asked providers to estimate the percentage of their Prop 36
clients that were homeless at treatment entry. The 84 providers that responded to this item
(97.6%) reported that, on average, 28.7% (± 28.5%) of their Prop 36 clients were homeless at
treatment entry. The percentages ranged from 0 to 100, with 10% being the most common
response. Additionally, of the 84 programs that responded to this item, 79 reported having
some portion of clients who were homeless.
In the UCLA Stakeholder survey, in response to the question “Were special strategies in
place for homeless Prop 36 offenders?” 60.4% (29) of the lead agencies endorsed “yes”.
Table 4.7 presents the methods that were used to address homelessness.

Table 4.7: Methods for Addressing Homeless
Method

N

%

Homeless received housing placement or assistance

26

89.7%

Homeless referred to residential treatment
Homeless were referred to treatment programs specializing in
treating homeless
Homelessness addressed with some other method such as referrals
to missions, shelters, or sober living environments

21

72.4%

8

27.6%

13

44.8%

Note: Percentages are based on the 29 lead agencies (60.4%) that reported using special strategies for homeless
clients.

Analysis of CADDS data across the first 5 years (July 1, 2001 to June 30, 2005) from the
clients who entered Prop 36 treatment gives some indication of the prevalence of
homelessness in the Prop 36 population by fiscal year. The CADDS treatment admission
form asks treatment providers “Is the person being assessed homeless?” While this item is
denoted as an “Optional Data Item” on the CADDS form, there are data for 22,457 clients
who entered treatment as part of Prop 36 in Year 1, 32,617 in Year 2, 33,761 in Year 3,
36,141 in Year 4, and 37,032 in Year 5 (see Table 4.8).

77

Table 4.8: Homelessness in Prop 36 Clients in Treatment
across the First 5 years of Prop 36
Year

N

%

Year 1 (FY 01/02)

2,153

9.6%

Year 2 (FY 02/03)

3,434

10.5%

Year 3 (FY 03/04)

3,258

9.7%

Year 4 (FY 04/05)

3,732

10.3%

Year 5 (FY 05/06)

3,787

10.2%

Note: Percentage reported is the number of “Yes” responses divided by the total
number that had data on the item “Is the person being assessed homeless?”.

Prevalence of Homelessness and Mental Illness in Proposition 36 Eligible Offenders
Client records that had data on the CADDS Homelessness item were matched to the DOJ
database to create a sample population for comparison purposes. Because the available
CADDS identifier is limited, it only contains first and last initial, sex, and date of birth (for
more information see Appendix 1.3), only a subset of records matched with the DOJ data.
For mental illness the best measure available was whether the participant is identified in the
DMH database as having received mental health services in the 12 months following the date
that they became eligible for Prop 36 participation. If they were not identified in the DMH
database, the assumption was that they were not suffering from a mental disorder in the 12
months following Prop 36 eligibility. The 12-month follow-up rule is the same standard
used to identify Prop 36 drug treatment following the eligible conviction. While this
approach is optimistic in its assumption that those with mental illness receive mental health
services, given the way data are collected, it is the most conservative method for making
comparisons. The alternative would be to use one of two items on CADDS, either the
Optional Data Item: “Has this person ever been diagnosed as also having chronic mental
illness?” or the Disability Impairment item that asks those completing the form to indicate up
to 3 disabilities that the client is impaired by, with Mental being one of the 8 options. The
primary concern with these items is that they are based on either the client’s self-report of
their mental health history or an estimation by the person completing the CADDS form, who
is likely not a mental health professional and thus does not have the training to make a
diagnosis.
In this sense, the appropriate terminology for the comparisons is those that are identified as
homeless and receiving mental health services (Homeless Mentally Ill) compared to those
that are identified as homeless not receiving mental health services (Homeless), those
identified as not homeless receiving mental health services (Mentally Ill), and those
identified as not homeless, not receiving mental health services (Neither). Table 4.9 presents
the N’s associated with each group of eligible offenders across the first 2 years of Prop 36.
These are the years being used as they have the most complete data and allow for a 30-month
follow-up period, which is consistent with the follow-up period of other outcomes presented
in this report.

78

Table 4.9: The Percentage of Prop 36 Eligible Offenders by Homeless and Mental
Health Status
Homeless
Year
Homeless
Mentally Ill
Neither
Mentally Ill
Year 1 (n = 12,521) 1.8%
11.5%
7.0%
79.6%
Year 2 (n = 13,335)

2.1%

11.3%

7.6%

79.0%

Note: It is unclear why the numbers present in the Homeless Mentally Ill and Mentally Ill groups for the third
year are so drastically lower than in the previous two years.

Analyses of the data presented in table 4.9 indicates that the proportion of individuals in the
Homeless Mentally Ill group is significantly higher than would be expected if the two events
(Homelessness and Mental Illness) occurred independently of each other in each year. In
other words, in proportion, a participant identified in the DMH database was more likely to
also be homeless, than an individual not identified in the DMH database.

Table 4.10: Age of Prop 36 Eligible Offenders by Homeless and Mental Health
Status
Homeless
Year
Homeless
Mentally Ill
Neither
Mentally Ill
Year 1
36.3 years
36.5 years
35.1 years
33.7 years
Year 2
35.5 years
36.4 years
35.1 years
33.7 years
Note: For all years the Homeless Mentally Ill and the Homeless groups were significantly older than the
Neither group. The Mentally Ill group was also significantly older than the Neither group.

Table 4.11: Race/Ethnicity of Prop 36 Eligible Offenders by Homeless and Mental
Health Status
Homeless
Year 1
Homeless
Mentally Ill
Neither
Mentally Ill
American Indian 0%
0.6%
0.7%
0.6%
Asian
0.9%
0.5%
0.9%
1.0%
Black
22.9%
21.3%
15.4%
12.2%
Hispanic
14.5%
20.9%
23.6%
31.3%
Pacific Islander
0%
0.8%
0.6%
1.3%
White
61.7%
55.1%
58.0%
52.4%
Other
0%
0.8%
0.8%
1.3%
Year 2
American Indian 0.7%
0.7%
0.5%
0.4%
Asian
0.7%
1.0%
0.5%
1.1%
Black
24.8%
17.4%
22.2%
12.0%
Hispanic
14.6%
22.2%
23.0%
32.1%
Pacific Islander
0%
0.7%
1.0%
1.3%
White
58.0%
54.4%
56.5%
51.8%
Other
1.1%
1.1%
0.8%
1.3%
Note: There were significantly differences among the Race/Ethnicity groups across the Homeless and Mentally
Ill variables. Across both years there significantly were more Whites and Blacks in the Homeless Mentally Ill
and Homeless groups than in the Mentally Ill and Neither groups. There were significantly fewer Hispanics in
the Homeless Mentally group than any other group.

79

Analyses of demographic data indicate significant age differences among the 4 groups (see
Table 4.10). Additionally, there were significant differences in the distribution of
Race/Ethnicity (see Table 4.11) and primary drug of choice (see Table 4.12).

Table 4.12: Primary Drug of Choice of Eligible Offenders
by Homeless and Mental Health Status
Homeless
Year 1
Homeless
Mentally Ill
Mentally Ill
Alcohol
18.6%
13.7%
10.2%

Neither

Cocaine/Crack

20.5%

18.4%

14.8%

11.3%

Heroin/Opiates

18.2%

15.7%

13.9%

11.9%

Marijuana

4.1%

6.4%

8.7%

11.5%

(Meth)amphetamine

36.8%

44.6%

50.5%

53.8%

Other

1.8%

1.2%

1.8%

1.8%

Alcohol

15.2%

12.9%

10.0%

8.9%

Cocaine/Crack

21.2%

19.3%

16.6%

11.0%

Heroin/Opiates

16.8%

14.7%

14.7%

10.9%

Marijuana

3.6%

6.5%

10.5%

11.8%

(Meth)amphetamine

41.2%

45.6%

46.9%

55.7%

Other

1.8%

0.9%

1.4%

1.7%

9.7%

Year 2

Note: There were significant differences among the 4 groups. Alcohol, Cocaine/Crack, and Heroin were more
prevalent in the Homeless Mentally Ill and Homeless groups than in the Mentally Ill and Neither groups across
the first 2 years. Methamphetamine was more prevalent in the Mentally Ill and Neither groups across the first 2
years.

Tables 4.13 and 4.14 present the Superordinate Diagnostic information by year and
Homeless status. Analyses indicated that there were no significant differences between the
two groups among the Superordinate Diagnostic categories across the two years.
Additionally, the data indicate that the ordering of the prevalence of the diagnostic
categories, while not as representative as the population as a whole, is similar across years.

Mental Illness and Homelessness in the Court Room
The UCLA Stakeholder Survey showed that each county handled Prop 36 clients with cooccurring mental illness and homelessness according to their own rules or available
resources. Of the 27 court administrators who completed the survey 77.8%, or 22, reported
that they assigned mental health services as needed. Additionally, 40.7%, or 11, reported
that they assigned some type of housing for homeless Prop 36 clients.

80

Table 4.13: Distribution of Diagnoses by Homeless Status among Year 1
Eligible Offenders
Superordinate Diagnostic Category
Mood Disorder
Substance
Schizophrenia/Psychotic
Adjustment
Anxiety
Deferred
V Code
Childhood
Mental
Impulse
Dissociative
Eating
Medical Condition
Personality
Cognitive

Homeless
38.3
26.9
17.6
7.0
5.3
2.2
0.0
0.4
0.4
0.4
0.0
0.0
0.0
0.4
0.9

Not Homeless
42.9
20.9
17.9
6.8
5.3
2.4
1.6
0.8
0.6
0.3
0.1
0.1
0.1
0.1
0.0

Table 4.14: Distribution of Diagnoses by Homeless Status among Year 2
Eligible Offenders
Superordinate Diagnostic Category
Mood Disorder
Substance
Schizophrenia/Psychotic
Adjustment
Anxiety
Deferred
V Code
Childhood
Impulse
Medical Condition
Mental
Personality
Abuse
Eating

Homeless
42.0
17.2
19.3
7.3
4.7
5.1
1.1
0.7
0.4
0.4
0.7
0.0
0.0
0.0

81

Not Homeless
38.7
20.8
17.1
6.7
6.4
5.1
1.8
1.4
0.6
0.5
0.4
0.4
0.1
0.1

UCLA also collected data from the public defenders assigned to handle Prop 36 cases. The
public defenders were asked whether their mentally ill and/or homeless clients gave different
reasons from those with stable housing and no mental disorders for declining Prop 36
participation. The majority of public defenders did not indicate differences in reasons for
refusal. However, among the 17.4% that reported differences (4 of 23 public defenders), the
different reasons were:
•

Lack of transportation to be able to get to mental health or drug treatment
and other numerous appointments associated with participating in Prop 36.

•

Concerns about being able to pay the fines associated with Prop 36 and the
difficulty of qualifying for Prop 36 without having an address.

•

Some drug treatment providers refused to accept clients with mental
illness or who were homeless.

•

The obligations of Prop 36 were too demanding.

These reasons are suggestive of the greater barriers to accessing treatment services that
homeless mentally ill face and to engaging this special population into treatment.

Treatment of the Mentally Ill in Prop 36
UCLA received responses to the UCLA Program Survey from 86 drug treatment providers
who held Prop 36 contracts at the time of data collection. One question on the survey asked
respondents (usually the drug use disorders treatment program director) to estimate the
percentage of their Prop 36 clients that were homeless and had a co-occurring mental
disorder. The treatment providers reported that, on average, 21.2% (ranging from 0 to 100)
of their Prop 36 clients who had a co-occurring mental disorder were homeless.
One aim of the survey was to determine how the drug use disorders treatment programs were
addressing the presence of mental disorders and homelessness among their clients. The
survey asked if the program employed mental health professionals (see Table 4.15).
For this analysis the NTP programs were removed as all would employ a doctor, likely a
psychiatrist, to oversee medication administration. Of the drug treatment programs that were
not primarily NTP programs (n = 84) that completed the UCLA Program Survey, 28.6%, or
24 programs, reported employing some combination of psychiatrists, psychologists, and
social workers. 9.4%, or 8, reported having at least one psychiatrist on staff, 7.1%, or 6,
reported having at least one psychologist on staff, and 21.4%, or 18, reported having at least
one social worker on staff.
The UCLA Program Survey also asked respondents: “What types of services have been
available to Prop 36 clients in the past fiscal year (7/1/06 to 6/30/07)?” The survey asked
respondents to indicate if 5 specific mental health services were available on site, by referral
through a cooperative agreement, or not available on site or by referral. Table 4.16 details
the number of programs that offered services on site.

82

Table 4.15: Distribution of Mental Health Professionals across
Treatment Programs
Program
Total Mental Health
Psychiatrist Psychologist Social Worker
Number
Professionals
1
2
2
2
1
2
3
3
2
2
4
1
1
5
1
1
6
1
1
2
7
2
2
8
2
1
3
9
4
4
10
1
1
2
11
1
1
12
1
1
13
1
1
14
1
1
15
1
1
16
1
1
17
2
1
3
6
18
1
1
19
1
1
1
3
20
3
3
21
1
1
22
2
2
23
2
2
24
1
1

Table 4.16: Methods for Addressing Mental Illness On Site
Services Available On Site

N

%

Mental Health Assessment and Diagnosis
Mental Health Counseling or Therapy
Mental Health Medication Services

26
33
19

30.9%
39.3%
22.6%

Dual Diagnosis Groups
Behavioral Interventions for Mental Health Problems

33
29

39.3%
34.5%

Note: N in this table represents the number of programs offering mental health services as part of their
treatment programs, meaning that they provide mental health and drug treatment services simultaneously.

There are two primary implications of these findings: 1) approximately 27% of the drug
treatment facilities holding Prop 36 contracts have the ability to offer mental health services
in an integrated fashion (i.e., receiving mental health and drug treatment at the same

83

treatment facility) and 2) programs that do not employ mental health professionals report
offering mental health services on site. The second implication can be explained in a number
of ways. Consulting data collected through focus groups, UCLA was able to determine that
drug treatment providers often came up with inventive ways to offer mental health services.
In cases where the county was the primary drug treatment provider, it was not unusual to
have the county mental health services located in the same or in nearby facilities. This colocation allowed providers to offer mental health services on site even when they were not
employing mental health professionals, especially if the Prop 36 participant had some
method to pay for the mental health services. However, UCLA also found instances where it
appeared that mental health services were being provided by individuals not adequately
trained to offer these services, such as dual diagnosis treatment groups run by certified
substance use disorders counselors. While it is clear that some form of mental healthcare
would be beneficial for the majority of Prop 36 clients, services offered by individuals not
adequately trained have the potential of causing harm to the clients, and thus violate industry,
state, and federal ethical and legal regulations, regardless of the intention of the provider.
Table 4.17 details the responses of the programs that reported not employing mental health
professionals, the remaining 73% of the 84 programs that completed UCLA’s Program
Survey, 6% (5) reported not offering nor having formal referrals for mental health
assessment or diagnosis, 9.5% (8) reported not offering nor having formal referrals for
mental health counseling, and 23.8% (20) reported not offering nor having formal referrals
for dual diagnosis groups.

Table 4.17: Programs Not Employing Mental Health Professionals
Programs Not Offering nor Referring for Mental Health Services
Mental Health Assessment and Diagnosis
Mental Health Counseling or Therapy
Mental Health Medication Services
Dual Diagnosis Groups
Behavioral Interventions for Mental Health Problems

N

%

5
8
19
20
12

5.8%
9.5%
22.1%
23.8 %
14.0%

It is interesting to note that, of the 24 clinics that are equipped to offer integrated mental and
drug use disorders treatment because they employed one of the three types of mental health
professionals listed above, 11 reported offering dual diagnosis treatment and all 11 reported
using a formal integration of these services called Integrated Dual Disorder Treatment
(IDDT). There were 8 programs that did not report employing mental health professionals
but reported offering IDDT, even though, according to the guidelines established regarding
IDDT, mental health professionals must be involved in treatment in order for the treatment to
be fully integrated, and thus formally considered IDDT.
Additionally, among the 86 treatment programs only 8.1% (7) reported conducting formal
psychodiagnostic assessment, 10.7% (9) indicated that they report client data to DMH.
Conversely 89.3% (75) of the programs report that they refer Prop 36 clients with a cooccurring mental illness to a mental health treatment provider, which means that even some

84

of the programs that indicated that they are dual diagnosis or IDDT programs also refer Prop
36 clients elsewhere for mental health services.

Treatment of the Homeless in Prop 36
Concerning the homeless, 89.5% (77) of the programs that responded to the survey reported
treating homeless persons. Table 4.18 details the methods the programs reported that they
used to address the homelessness of their clients.
Table 4.18: Housing Services offered by Prop 36 Treatment Providers
Service
N
%
Place homeless in residential treatment
46
54.8%
Attempt to find housing through AB2034
23
27.4%
Other stable housing
67
79.8%
Other services (sober-living beds, referrals to shelters)
31
36.9%
Note: Percentages are based on the77 of the programs (89.5%) that reported treating homeless clients.

In sum, the survey findings show that there is considerable capability to provide services to
this population, although the strategies used and professional mental health training vary
considerably across providers. There remain significant gaps across the providers in the
provision of services that have been defined as “Best Practices” for this population.

Outcomes for the Homeless with Mental Illness among Prop 36 Eligible Offenders
In order to determine the effectiveness of Prop 36 for this special population, UCLA
compared the homeless with co-occurring disorders identified and matched across the three
databases (DOJ, DMH, and CADDS) to those that reported not being homeless in CADDS
and those who where not located in the DMH administrative database. Those that had any
data on the Homelessness item on CADDS in the matched database were 1,407 (or 3.6% of
the entire Prop 36 eligible population) in year 1 and 2,194 (or 4.8% of the entire Prop 36
eligible population) in year 2. Given these percentages the obvious caveat is that the
available data are not necessarily a random or representative sample of homelessness in the
Prop 36 population, so caution should be used when interpreting them. The same holds true
for those identified in the DMH administrative database as having a co-occurring mental
disorder. In these analyses any one not present in the DMH database is included in the
neither group. This means that the true nature of the comparison is those that received
mental health services versus those that did not or did not receive them through a DMH
provider. Finally, given the amount of missing data, comparisons of treatment completion
and treatment duration are based on the data available.
Treatment Placement
The first step in the comparisons was to determine if there were significant differences
between the groups in the type of treatment they were placed in (see Figures 4.1 and 4.2).

85

Figure 4.1 Year 1 Treatment Modality by Homelessness and
Mental Health Status (n = 12,521)
100.0%
80.0%

70.5%

76.4%

60.0%
44.3%

40.0%

37.4%

34.8%

31.9%
16.1%

14.5%12.3%

20.0%

3.9% 2.7%

4.8% 4.0% 1.6% 1.8%

10.3%

0.0%

Outpatient Drug Free

Detoxification

Homeless Mentally Ill

Residential < 30 Days Residential > 30 Days

Homeless

Mentally Ill

Neither

Note: Methadone detoxification and methadone maintenance are not included as less than 3% of the clients
in any group received them and the difference in placement was not significantly different across groups.

Figure 4.2
Year 2 Treatment Modality by Homelessness and
Mental Health Status (n = 13,335)
100.0%
80.0%
60.0%
40.0%

80.4%
75.0%

45.4%
36.5%

42.0%
31.7%
15.0%14.6%

20.0%

3.8% 2.7%

13.9%
10.2%

3.6% 3.9% 1.7% 1.4%

0.0%

Outpatient Drug Free

Detoxification

Homeless Mentally Ill

Residential < 30
Days
Homeless

Mentally Ill

Residential > 30
Days
Neither

Note: Methadone detoxification and methadone maintenance are not included as less than 3% of the clients
in any group received them and the difference in placement was not significantly different across groups.

The pattern of placement in different treatment modalities was the same across all three
years. The Homeless Mentally Ill and the Homeless were significantly more likely to be
placed in Detoxification, Residential < 30 Days, and Residential > 30 days than the Mentally

86

Ill and Neither groups. The Mentally Ill and Neither groups were more likely to be placed in
Outpatient Drug Free.
Treatment Completion
One of the primary outcome measures for the evaluation of Prop 36 is treatment completion.
Using available data treatment completion for those clients with homelessness data are
presented in tables 4.3 and 4.4.

Figure 4.3
Year 1 Drug Treatment Completion by Homelessness and
Mental Health Status (n = 10,028)
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%

59.0%

52.9%

57.5%

50.1%

39.1%

36.5%

35.1%

33.2%

10.6%

10.8%

7.7%

7.4%

Homeless Mentally Ill

Homeless

Mentally Ill

Neither

Incomplete Unsatisfactory

Incomplete Satisfactory

Complete

Note: Percentage based on those clients that had a discharge status. Clients with no discharge status were
treated as missing data.

Figure 4.4
Year 2 Drug Treatment Completion by Homelessness and
Mental Health Status (n = 11, 321)
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%

63.0%

57.1%

57.1%

51.2%
39.7%

33.3%

35.1%

29.4%

9.5%

9.1%

7.5%

7.2%

Homeless Mentally
Ill

Homeless

Mentally Ill

Neither

Incomplete Unsatisfactory

Incomplete Satisfactory

Complete

Note: Percentage based on those clients that had a discharge status. Clients with no discharge status were
treated as missing data.

Treatment Duration
In addition to completion information, using CADDS data, the length of time in treatment
can be calculated for individuals that have both an intake and a discharge date. As noted in
other parts of the report this is an estimate for only those people that have a discharge status.

87

It is difficult to interpret the data given the amount of missing discharge data, however, using
available data may give some indication of how long this special population remains in
treatment and how that compares to other populations. Additionally, this variable does not
account for the amount of actual time in treatment, just the time between intake and
discharge. Table 4.19 presents data on the length of time spent in treatment by
Homelessness and Mental Health status.

Table 4.19: Days in Drug Treatment by Homeless and Mental Health Status
Homeless Mentally Ill Homeless

Mentally Ill

Neither

Year 1 (n = 9659)

78.4 (117.2)

100.7 (134.2)

139.4 (145.0) 142.2 (147.9)

Year 2 (n = 10,927)

80.1 (126.7)

101.6 (129.9)

131.0 (145.8) 145.6 (152.6)

Note: Data present are Mean and (Standard Deviation) based on those that had discharge data. All differences
are significant except the difference between the Year 1 Mentally Ill and Neither groups.

It is interesting to note that, across all three years, the Homeless Mentally Ill spent the fewest
days in treatment followed by the Homeless, then the Mentally Ill, and finally the Neither
group. As previously noted the homeless mentally ill are particularly difficult to retain in
treatment for a number of reasons, such as unstable housing and a higher likelihood of rearrest while in treatment.
Re-offending
In addition to drug treatment outcome data, DOJ data was used to compare re-offending
among the Homeless Mentally Ill, the Homeless, the Mentally Ill, and the Neither groups.
Figures 4.5 and 4.6 present the percentage from each group that was arrested at least one
time in the 30 months following their Prop 36 eligible conviction. Only property and violent
crime data are presented as other crime types were not prevalent enough across all three
years to warrant analysis.

Figure 4.5
Year 1 Re-Offending among Prop 36 Eligible Offenders
(n = 12,502)
70.0%

63.0%

57.6%

56.1%

60.0%

53.3%

50.0%
40.0%
30.0%

20.7%

10.0%

17.2%

17.1%

20.0%
5.7%

14.4%
5.0%

3.2%

3.8%

0.0%

Homeless Mentally
Ill

Homeless

Drug Crime

Mentally Ill

Property Crime

Neither

Violent Crime

Note: Shows the percentages of the total participants per group re-arrested at least once for each type of crime.

88

Figure 4.6
Year 2 Re-Offending among Prop 36 Eligible Offenders
(n = 13,335)

70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%

63.5%

57.4%

56.3%

21.2%
8.8%

Homeless Mentally
Ill

18.2%

16.8%
4.1%

Homeless
Drug Crime

49.4%

6.4%

Mentally Ill

Property Crime

14.2%

3.8%

Neither

Violent Crime

Note: Shows the percentages of the total participants per group re-arrested at least once for each type of crime.

Analysis of the new arrest data indicated that, across both years, the Homeless Mentally Ill
and the Homeless groups were more likely than the Mentally Ill and Neither groups to be
arrested for all types of crime (property, violent, or drug) in the 30 months following the
conviction that made them eligible for Prop 36 participation. Additionally, the Mentally Ill
group was more likely to have a new arrest for a drug, property, or violent crime than the
Neither group across all three years. These results are consistent with previous research
presented at the beginning of this chapter that indicates that homeless mentally ill clients are
likely to have worse treatment outcomes when compared to the homeless, the mentally ill,
and those individuals that have stable housing and no mental illness.

Barriers to Treating the Homeless with Co-Occurring Disorders in Prop 36
UCLA collected data through focus groups conducted across the state with various groups of
stakeholders (i.e., lead agencies, probation, treatment providers). As part of these focus
groups, UCLA asked the stakeholders to identify barriers to treating special populations such
as the homeless with co-occurring mental disorders. These stakeholders noted that the
primary barrier to treating this population is that mental health services are not considered
part of drug treatment and, as such, Prop 36 money cannot be used to explicitly pay for
mental health services. Other barriers that were identified include that transience that is
inherent in homelessness, which makes it difficult to track homeless clients. Additionally,
the homeless and those with co-occurring disorders require more treatment and use more
resources than those with stable living arrangements or who do not have mental illness.
Some lead agencies noted the need for more transient housing services. Others stated that
the requirements of Prop 36 are too cumbersome for the homeless and those with mental
illness.
Other stakeholders were able to identify strategies that they used to deal with these barriers.
In counties where the primary drug treatment is provided by the county, the county only
hires mental health professionals so that they could provide integrated services. Other
counties have contracted with local shelters to be able to provide housing to the homeless
while they are in treatment. In other counties courts have set up dedicated co-occurring

89

disorders court calendars and have dedicated staff to deal with the needs of this special
population. Some stakeholders were able to secure resources for transient housing by joining
their county’s continuum of care board. This opened the way to get money for motel
vouchers, and that qualified them to apply for money through the state Housing and
Community Development Department. The county used this grant to open their own
transitional housing.

Co-Occurring Disorders and Homelessness in the Offender Treatment Program
ADP outlined a list of goals and strategies for counties to focus on for OTP funding which
was largely based on recommendations from UCLA’s ongoing evaluation of Prop 36. One
was to develop treatment services that are needed but not available.
Thirty-nine counties submitted applications for OTP funding. UCLA coded the applications
which detailed how the requested funds would be used. Approximately 23% of the counties
(9) specified establishing treatment groups designed to serve those with co-occurring
disorders. Approximately 10% (4) indicated that they would increase transitional housing.
Approximately 5% (2) indicated that they would like to add residential slots for clients with
co-occurring disorders. This indicates that many counties saw the need to increase services
available to those with co-occurring disorders and unstable housing in Prop 36. What is
unclear from the OTP process is whether those that did not indicate increasing services to
this special population thought they had sufficient resources for this population, did not offer
services, or did not have enough OTP funding to meet all of the goals outlined in OTP.
Additionally, there is, as of yet, no indication that the services were actually implemented in
the counties that sought funds to do so.

Recommendations
Identification of Those in Need
There are currently no standards for the initial assessment process across counties. Most
counties conduct an assessment that includes some form of the ASI or the ASAM-PPC.
Others use measures that they themselves have developed. Additionally, counties are not
required to report the assessment results to any centralized database. CalOMS does not
provide sufficient information to make accurate estimates of the prevalence of mental health
disorders in the AOD treatment population, nor the ability to distinguish between less and
more severe mental disorders. This makes identifying and tracking the homeless with cooccurring mental disorders in treatment almost impossible.
UCLA is recommending that a standardized assessment be either a) adopted or b) developed
and then implemented statewide. This assessment tool should accurately identify those who
are homeless or in danger of becoming homeless and those that have mental illness that is
significantly affecting their functioning according to DSM-IV-TR criteria, in addition to the
other areas that are typically assessed as part of the Prop 36 assessment. A standardized
training should also be developed so that each person charged with assessing Prop 36 clients
receives an appropriate level of training to correctly and accurately conduct the assessment.
Additionally, UCLA is recommending that counties report assessment results in a
standardized form to a database maintained by ADP, either as part of CalOMS or through a
separate database, for research and quality assurance purposes.

90

Proper Treatment Placement for Individuals with Co-Occurring Disorders
System-level efforts to improve treatment for clients with co-occurring disorders have
included the development of a conceptual framework for placing clients with co-occurring
disorders in the level of treatment most suited to the severity of combined disorders, as
exemplified by the “quadrant” model (National Association of State Mental Health Program
Directors and National Association of State Alcohol/Drug Abuse Directors, 1998; Burnam &
Watkins, 2006; Pincus et al., 2007). This model suggests that individuals who are “high” in
severity on both dimensions of substance use and mental disorders require treatment in highintensity settings, such as residential treatment, whereas those low in severity in both or
either dimension can be effectively treatment in specialized mental health or substance abuse
treatment programs that have cross-linkages with programs in the other treatment sector (see
Figure 4.7). The feasibility of using this model to classify clients with co-occurring
disorders into the appropriate level of care was recently supported in a study using Medicaid
claims data from 6 states (McGovern et al., 2007).

Substance Use Disorders

High
Severity

Figure 4.7
Level of Care Quadrants
Category IV
Disorder Severity:
Mental Disorders more severe
Substance Use Disorders more severe

Category III
Disorder Severity:
Mental Disorders less severe
Substance Use Disorders more
severe

Locus of Care:
Substance Use Treatment System

Locus of Care:
Intensive Integrated Treatment,
usually in a residential placement

Category I
Disorder Severity:
Mental Disorders less severe
Substance Use Disorders less
severe

Category II
Disorder Severity:
Mental Disorders more severe
Substance Use Disorders less
severe

Locus of Care:
Primary Health Care Settings

Locus of Care:
Mental Health System

Low
Severity

High
Severity

Mental Disorder

91

Community-based residential programs may be particularly appropriate for providing a
broad range of integrated services for homeless individuals who are “high” on both mental
health and substance use severity. These programs include mental health treatment,
substance abuse interventions, transitional housing, life and social skills, and other supports.
A recent review of 10 controlled studies suggests that greater levels of integration of
substance abuse and mental health services are more effective than less integration for
treating co-occurring disorders (Brunette et al., 2004). Further, when mental health services
are located on-site in residential programs, individuals are more likely to obtain these
services and to have lower drug use and better mental health status at 6 months following
treatment (Grella & Stein, 2006). The therapeutic community model of treatment has been
adapted for individuals with co-occurring disorders (De Leon et al., 2000), and is the most
frequently used treatment approach within prison settings (Peters et al., 2004).
Integrated Dual Disorders Treatment
Several other treatment approaches have been adapted for use with homeless individuals
with co-occurring disorders who come into contact with the criminal justice system, with the
goals of improving community functioning and preventing jail detention and recidivism
(Chandler & Spicer, 2006; Drake et al., 2006). There is increasing emphasis on utilizing
evidence-based practices for this population (Chandler et al., 2004). These include assertive
community treatment, intensive case management, and integrated dual disorders treatment
(IDDT). One controlled trial compared Integrated Assertive Community Treatment,
Assertive Community Treatment only, and standard care among homeless clients with cooccurring disorders. The study found that although there were no significant differences
among groups in substance use or psychiatric symptoms, subjects in the two experimental
conditions reported more days in stable housing over a 24 month follow-up period, compared
with those in standard care (Morse et al., 2006). Other promising approaches have focused
on improving transition planning at the time of leaving jail or paroling from prison into the
community (Osher et al., 2003).
IDDT is an evidence-based practice for the treatment of co-occurring disorders that the
Substance Abuse and Mental Health Services Administration (SAMSHA) currently
recommends as the preferred treatment for individuals diagnosed with co-occurring disorders
(SAMSHA, 2003). The majority of the data published to date supports improved treatment
outcomes for those receiving IDDT compared to care-as-usual, such as parallel or serial
treatments of the mental and substance use disorder (i.e., Boyle & Kroon, 2006; James et al.,
2004; and Mangrum et al., 2006). Additionally, the IDDT approach includes a performance
management component that allows for oversight of these programs.
UCLA recommends that each county be able to offer IDDT to those who meet diagnostic
criteria for co-occurring disorders. While UCLA realizes that implementing an IDDT
approach, even in one treatment setting, will be resource intensive upfront, however, the long
term improvement in outcomes expected from adopting this approach would offset this
initial resource investment. First steps towards adopting an IDDT approach can be
accomplished in a number of ways, one would be to award Prop 36 contracts to mental
health facilities that have or are willing to employ certified drug treatment counselors.
Conversely, Prop 36 contracts could mandate that drug treatment facilities employ licensed
mental health professionals as part of their regular full- or part-time staff. In addition Prop

92

63 grants could be awarded to drug treatment facilities that employ mental health
professionals. In addition to the ability to offer mental health services, these staff can assist
with finding housing placements and helping this special population get registered for
additional public assistance programs.
Programs could also design training programs to maximize their use of mental health
trainees. This would require hiring one mental health professional from a particular domain
(i.e., clinical psychology, clinical social work, or psychiatry) and then hiring trainees to be
supervised by the licensed professional. This maximizes investment in mental health
services while training professionals to continue serving this special population.
Alternate Funding Sources
In November 2004, California voters passed Prop 63, the Mental Health Services Act
(MHSA), with 53.4% of the vote. Prop 63 was designed to provide funds to counties to
expand services and develop innovative programs and integrated service plans for mentally
ill children, adults and seniors (Scheffler & Adams, 2005). The new law also established the
Mental Health Services Oversight and Accountability Commission (MHSOAC). This
commission recently released a report on co-occurring disorders that also listed specific
recommendations (MHSOAC, 2007).
One of the primary recommendations from this report is to take a “Whatever it Takes”
approach to funding and providing treatment refers to funding for a “wide array of clinical
and supportive services beyond mental health care, notably including such things as housing
and treatment for co-occurring [disorders]”. The commission also noted that IDDT was the
exception rather than the rule in California, but that integrated care is likely the best
treatment setting for individuals with co-occurring disorders and that there are limited public
and private funding sources for such integrated care. One of the primary recommendations
from this commission was that “Public and private health plans which have programs that are
funded by the Mental Health Services Act should be required to ensure integrated mental
health and substance abuse services are available for all clients who need them”.
Substance use disorder treatment programs that employee mental health professionals are
likely to be one of the best places to start co-locating these services, as they already have in
place the personnel necessary to offer integrated treatment. Alternatively, mental health
treatment facilities that employ certified substance use disorders treatment personnel are
equally equipped. It then becomes an issue of training and funding. As noted, the materials
needed for implementation of IDDT are available from SAMSHA at no cost.
In addition to awarding Prop 36 treatment contracts to mental health treatment facilities that
employ drug use disorders counselors, as noted above, UCLA is recommending that Prop 63
funds should be awarded to drug use disorders treatment facilities that employ mental health
professionals so that they can begin using an IDDT approach. This would allow the best use
of available funds from both sources to create and implement the “Whatever it Takes”
approach to treating the Homeless Mentally Ill.
As UCLA learned from the focus group data, some counties are already moving towards this
approach, as they have identified the need for Prop 36 dedicated Co-Occurring Disorders

93

Courts (Santa Barbara and Los Angeles Counties as examples) and Whatever It Takes Courts
(Orange County as an example). Currently these courts are in need of funds to continue and
expand these programs and it seems that an integration of Prop 36 and Prop 63 funds may be
the best means for accomplishing this goal.
Data Collection
As part of this evaluation UCLA was asked to evaluate the performance and outcomes of
Prop 36 clients that report being homeless and having a mental disorder. While this
population is of interest and the question is meaningful, it was difficult to answer given the
nature of the data available. As such, UCLA is recommending that ADP and county
stakeholders work together to develop better data collection so that special populations can
be easily identified and studied across multiple administrative databases.

Conclusions
Prop 36 clients who are homeless and have co-occurring mental and drug use disorders
represent a special population that are often difficult to track, study, and treat. Research
indicates that they are at increased risk for not completing treatment, recidivating, and
services provided to them are often more intense and, thus, more costly. Identifying clients
who meet these criteria early in the Prop 36 process, such as during assessment, may lead to
better outcomes. Additionally, these clients will likely benefit from placement in treatment
programs that are better suited to meet their needs, such as programs that offer integrated
mental health and drug use disorder treatment services.
Finally, California currently has two separate agencies with two separate means for funding
drug and mental health treatments for criminal justice offenders meeting Prop 36 eligibility
requirements, ADP with Prop 36, and the DMH with Prop 63. This separation seems to be
an impediment to integrating care for the homeless mentally ill drug users in Prop 36.
UCLA is recommending close collaboration between the Department of Alcohol and Drug
Problems and the Department of Mental Health to provide integrated services for those with
co-occurring disorders. Collaboration will likely lead to creative integration of the two
funding sources, Prop 36 and Prop 63, to maximize the value of the dollars provided by both
funding sources to best treat homeless individuals with co-occurring mental and drug use
disorders. For example, this collaboration could lead to awarding Prop 36 contracts and Prop
63 grants to IDDT facilities. The first step in this process may be the creation of “Whatever
It Takes” courts across the state. Specialized courts that are staffed by individuals with
specialized training in working with the homeless mentally ill may lead to the best outcomes
for this special population.

References
American Psychiatric Association (2000). Diagnostic and Statistical Manual of Mental
Disorders, 4th edition, Text Revision (DSM-IV-TR). Washington, DC: American
Psychiatric Press, Inc.
Boyle, P., & Kroon, H. (2006). Integrated dual disorder treatment: comparing facilitators
and challenges of implementation for Ohio and the Netherlands. International
Journal of Mental Health, 35, 70-88.

94

Brunette, M.F., & Mueser, K.T. (2006). Psychosocial interventions or the long-term
management of patients with severe mental illness and co-occurring substance use
disorder. Journal of Clinical Psychiatry, 67, Suppl 7, 10.17.
Brunette, M.F., Mueser, K.T., & Drake, R.E. (2004). A review of research on residential
programs for people with severe mental illness and co-occurring substance use
disorders. Drug and Alcohol Review, 23, 471-481.
Burnam, M.A., & Watkins, K.E. (2006). Substance abuse with mental disorders:
Specialized public systems and integrated care. Health Affairs, 25, 648-658.
Caton, C.L.M. (1990). Homeless in America. New York: Oxford University Press.
Chandler, D.W., & Spicer, G. (2006). Integrated treatment for jail recidivists with cooccurring psychiatric and substance use disorders. Community Mental Health
Journal, 42, 405-425.
Chandler, R.K., Peters, R.H., Field, G., & Juliano-Bult, D. (2004). Challenges in
implementing evidence-based treatment practices for co-occurring disorders in the
criminal justice system. Behavioral Sciences & the Law: Co-Occurring Disorders
and the Criminal Justice System, 22, 431-448.
Conway, K.P., et al., (2006). Lifetime comorbidity of DSM-IV mood and anxiety
disorders and specific drug use disorders: Results from the National Epidemiologic
Survey on Alcohol and Related Conditions. Journal of Clinical Psychiatry, 158,
420-426.
De Leon, G., Sacks, S., Staines, G., & McKendrick, K. (2000). Modified therapeutic
community for homeless mentally ill chemical abusers: Treatment outcomes.
American Journal of Drug and Alcohol Abuse, 26, 461-480.
Drake, R.E. et al. (2001). Implementing dual diagnosis services for clients with severe
mental illness. Psychiatric Services, 52, 469-476.
Drake, R.E., et al. (2006). The challenge of treating forensic dual diagnosis clients:
Comment on "integrated treatment for jail recidivists with co-occurring psychiatric
and substance use disorders". Community Mental Health Journal, 42, 427-432.
Gonzalez, G., & Rosenheck, R.A. (2002). Outcomes and service use among homeless
persons with serious mental illness and substance abuse. Psychiatric Services, 53,
437-446.
Grella, C.E., & Stein, J.A. (2006). Impact of program services on treatment outcomes of
patients with comorbid mental and substance use disorders. Psychiatric Services, 57,
1007-1015.
Hartwell, S. (2004). Triple stigma: Persons with mental illness and substance abuse
problems in the criminal justice system. Criminal Justice Policy Review, 15, 84-99.
James, W., Preston, N.J., Koh, G., Spencer, C., Kisely, S.R., & Castle, D.J. (2004). A
group intervention which assists patients with dual diagnosis reduce their drug use: a
randomized controlled trial. Psychological Medicine, 34, 983-990.
Kessler, R.C. (2004). The epidemiology of dual diagnosis. Journal of Biological
Psychiatry, 56, 730-737.

95

Kushel, M.B., Hahn, J.A., Evans, J.L., Bangsberg, M.D., & Moss, A.R. (2005).
Revolving doors: Imprisonment among the homeless and marginally housed
population. American Journal of Public Health, 95, 1747-1752.
Lamberti, J.S., Weisman, R.L., Schwarzkopf, S.B., Price, N., Ashton, R.M., &
Trompeter, J. (2001). The mentally ill in jails and prisons: Towards an integrated
model of prevention. Psychiatric Quarterly, 72, 63-77.
Mangrum, L., Spence, R., & Lopez, M. (2006). Integrated versus parallel treatment of
co-occurring psychiatric and substance use disorders. Journal of Substance Abuse
Treatment, 30, 79-84.
McGovern, M.P., Clark, R.E., & Samnaliev, M. (2007). Co-occurring psychiatric and
substance use disorders: A multistate feasibility study of the quadrant model.
Psychiatric Services, 58, 949-954.
McGovern, M.P., Xie, H., Segal, S.R., Siembab, L., & Drake, R.E. (2006). Addiction
treatment services and co-occurring disorders: Prevalence estimates, treatment
practices, and barriers. Journal of Substance Abuse Treatment, 31, 267-275.
McNiel, D.E., Binder, R.L., & Robinson, J.C. (2005). Incarceration associated with
homelessness, mental disorder, and co-occurring substance abuse. Psychiatric
Services, 56, 840-846.
Medi-Cal Eligibility Manual (2007). California Medi-Cal Eligibility Branch. Accessed
at: http://www.dhs.ca.gov/mcs/mcpd/meb/MediCalEligibilityProceduresManual/default.htm
Mental Health Services Oversight and Accountability Commission (2007). Mental
Health Services Oversight and Accountability Commission Report on Co-Occurring
Disorders. Accessed at: http://www.dmh.ca.gov/MHSOAC/docs/CoOccurringDisorders.pdf.
Monahan J., Steadman H.J., Robbins P.C., Appelbaum P., Banks S., Grisso T., Heilbrun
K., Mulvey E.P., Roth L., Silver E. (2005). An actuarial model of violence risk
assessment for persons with mental disorders. Psychiatric Services, 56, 810-815.
Morse, G.A., et al. (2006). Treating homeless clients with severe mental illness and
substance use disorders: Costs and outcomes. Community Mental Health Journal,
42, 377-404.
Mueser, K.T.; Noordsy, D.L.; Drake, R.E.; Fox, L. (2003). Integrated Treatment for
Dual Disorders: A Guide to Effective Practice. New York. Guilford Press. 402-405.
NASMHD &NASADAD (1998). National dialogue on co-occurring mental health and
substance abuse disorders. Washington, DC: Center for Mental Health Services and
the Center for Substance Abuse Treatment of the Substance Abuse and Mental
Health Services Administration (SAMHSA).
National Comorbidity Study. Accessed at:
http://www.hcp.med.harvard.edu/ncs/index.php.
National Comorbidity Study Replication. Accessed at:
http://www.hcp.med.harvard.edu/ncs/index.php.

96

National Epidemiologic Survey on Alcohol and Related Conditions (NESARC).
Accessed at: http://niaaa.census.gov/.
Osher, F.C., & Dixon, L.B. (1996). Housing for persons with co-occurring mental and
addictive disorders. New directions for Mental Health Services, 70, 53-64.
Osher, F.C., Steadman, H.J., & Barr, H. (2003). A best practice approach to community
reentry from jails for inmates with co-occurring disorders: The APIC model. Crime
& Delinquency, 49, 79-96.
Peters, R.H., LeVasseur, M.E., & Chandler, R.K. (2004). Correctional treatment for cooccurring disorders: Results of a national survey. Behavioral Sciences & the Law,
22, 563-584.
Pincus, H.A., Watkins, K., Vilamovska, A., & Keyser, D. (2007). Models of care for cooccurring disorders: Final report to the Substance Abuse and Mental Health
Administration: Center for Substance Abuse Treatment. Santa Monica, CA: RAND
Corporation.
Regier, D.A., Farmer, M.E., Rae, D., Locke, B.Z., Keith, S.J., Judd, L.L., Goodwin, F.K.
(1990). Comorbidity of Mental Disorders with Alcohol and Other Drug Abuse.
Journal of the American Medical Association; 264, 2511-2518.
Rossi, P.H. (1989). Down and out in America: The origins of homelessness. Chicago:
University of Chicago Press.
Scheffler, R.M., & Adams, N. (2005). Millionaires and mental health: Proposition 63 in
California. Health Affairs. Web Exclusives: W5-212-W5-224. Accessed at:
http://content.healthaffairs.org/cgi/content/abstract/hlthaff.w5.212v1.
Substance Abuse and Mental Health Administration (SAMSHA; 2003). Co-Occurring
disorders: Integrated Dual Disorders Treatment. Accessed at:
http://mentalhealth.samhsa.gov/cmhs/communitysupport/toolkits/cooccurring/.
Watkins, K.E., Hunter, S.B., Wenzel, S.L., Tu, W., Paddock, S.M., Griffin, A., Ebener, P.
(2004). Prevalence and characteristics of clients with co-occurring disorders in
outpatient substance abuse treatment. American Journal of Drug and Alcohol Abuse,
30, 749-764.
Wenzel, S.L., Burnam, A.M, Koegel, P., Morton, S.C., Miu, A., Jinnett, K.J., & Sullivan,
G.J. (2001). Access to inpatient or residential substance abuse treatment among
homeless adults with alcohol or other drug use disorders. Medical Care, 39, 11581169.
Zuvekas, S.H., & Hill, S.C. (2000). Income and employment among homeless people:
the role of mental health, health and substance abuse. The Journal of Mental Health
Policy and Economics, 3, 153-163.

97

98

PROMISING PRACTICES
Chapter 5: Emerging Promising Practices in Proposition 36
Darren Urada, Ph.D.
Implementation of the recommendations included in previous Prop 36 evaluations has
been facilitated by the Offender Treatment Program (OTP). Responses to a UCLA survey
indicated that 37.5% of the OTP counties had not fully implemented their OTP activities
at the end of the 2006-2007 fiscal year. Nevertheless, the results of targeted treatment
expansions were readily detectable. In counties that used OTP funds to target expansion
in narcotic replacement therapy (NRT), the number of Prop 36 clients receiving NRT rose
97.3% over the number receiving NRT in the prior year. In counties that used funds to
expand residential treatment, the number of Prop 36 clients receiving this form of
treatment increased 8.1% over the prior year.
In addition to improvements being facilitated by OTP funding, several additional
innovations may improve program performance and client outcomes in Proposition 36.
Based on the research literature and interviews with stakeholders, a number of promising
practices were selected for further study. The chapters that follow in this section of the
report focus on five practices that appear to have strong potential to improve Prop 36
implementation. These include practices already being facilitated by OTP such as
narcotic replacement therapy, residential treatment, and drug testing and sanctions, as
well as employment assistance and process improvement. For short term purposes,
UCLA recommends making OTP or other funds available to facilitate implementation of
these practices. However, more stable, longer term sources of funding may be needed to
sustain improvements in these areas.
One goal of this evaluation was to review a number of evidence-based strategies that could
be used to reduce no shows into Prop 36 treatment, to retain offenders that are placed in
treatment, and to improve program outcomes. A number of recommendations have been
made as part of UCLA’s previous 2001-2006 evaluation of Prop 36. These
recommendations were:
•

Funding should be allocated to ensure greater availability of favorable drugtreatment options. Residential treatment should be available for those with
the most severe drug abuse as determined by a standardized assessment.
Narcotic Treatment Programs (e.g., methadone maintenance) should be used
as a first line intervention for those Prop 36 treatment clients with heroin or
other opiate use problems.

•

Practices associated with better Prop 36 show rates should be pursued,
including locating assessment units in or near the court, performing
assessments in a single visit, allowing walk-in assessments without
appointments, and incorporating procedures used in drug courts (e.g. a court
calendar dedicated to drug offenders, dialog between the judge and offender,
close supervision, and collaboration involving judge, prosecutor, defense
attorney, and treatment provider). Evidence-based practices established by
99

existing research should also be incorporated wherever possible, and
financial incentives should be considered for counties and providers for
instituting these practices or for otherwise demonstrating more success on
objective measures such as reduced time from Prop 36 conviction to
treatment entry.
•

Explore handling offenders with high rates of prior convictions differently.
This could include placement into more-controlled treatment settings (e.g.,
residential treatment), more intensive supervision, or drug court referral.

•

Collaboration and coordination among court, probation, parole, and treatment
systems should continue to be improved with the goal of admitting offenders
into appropriate treatment in the shortest possible time while maintaining
appropriate levels of oversight and supervision.

•

Drug testing information should be considered to provide an objective basis
for delivery of additional services or for a program of graduated sanctions for
offenders who are not complying with Prop 36 requirements.

•

A concerted, collaborative effort should be made to streamline access to and
use of state data for authorized evaluation studies. Efforts to improve the
quality of data sources such as the SACPA Reporting Information System are
also important.

•

Further policy-relevant sub-studies should be conducted to address issues that
remain, including research on barriers to success and potential
implementation improvements for Hispanics, parolees, offenders with cooccurring mental disorders, women, pregnant women and women with
children. Research is also recommended to investigate the net effect of Prop
36 on crime among the broader population of both drug offenders and nondrug offenders.

Recommendations and the Offender Treatment Program
Many of the recommendations listed above subsequently became goals and strategies in the
Offender Treatment Program (OTP), which was created by Assembly Bill 1808 for the
purpose of improving performance and outcomes in Prop 36. For Fiscal Year 2006-2007,
$25 million was allocated to this program to provide funding for Prop 36 improvements.
Funding was awarded to counties if they met certain eligibility requirements and if the
counties proposed strategies consistent with a list of recommendations compiled by ADP and
others. In all, 39 counties received funds from OTP in 2006-2007.
Performance
UCLA examined and coded the 39 OTP applications to categorize the types of goals and
activities each county intended to pursue. The most common goals and strategies proposed
are shown in Table 5.1.

100

Table 5.1: Offender Treatment Program Goals and
Strategies in County Applications (n=39) for OTP Funding
Goals and Strategies

Counties

Increase utilization of residential treatment services

23

Reduce treatment delays

23

Enhance criminal justice supervision

22

Expand access and treatment capacity

22

Expand residential bed capacity

19

Increase access to culturally relevant services for special populations

18

Intensify/add services as indicated by drug tests

18

Increase treatment oversight and supervision

16

Add probation officer(s)

15

Improve criminal justice, treatment coordination to reduce wait time

15

Increase outpatient services/expand outpatient capacity

14

Provide transportation (e.g., bus tickets, taxi vouchers)

14

Improve location of services to ensure access to all people

13

Increase narcotic replacement therapy

13

Use intermediate sanctions of graduating severity (not jail at this time)

13

Add counselor(s)

12

Develop sober living arrangements tied to outpatient treatment

12

Greater utilization of probation and program drug testing results

11

Conduct assessment & treatment in a single visit

9

Establish treatment groups to serve those with co-existing disorders

9

Add comprehensive case management/add case manager(s)

8

Increase drug testing

8

Increase other treatment services

8

Parenting/life skills/mental health/domestic violence/anger mgmt counseling

8

Provide psychiatric medications, support

7

Adopt drug court model

6

Expand sober living

6

Implement/continue/expand Matrix Model groups

6

Maintain dedicated court calendar

6

Utilize additional screening tools

6

101

In August 2007, UCLA sent surveys to all Prop 36 county lead agencies (for methods, see
Appendix A). Of the 39 counties that received OTP funds, 32 responded to the survey
(82.1%). To determine the status of these OTP efforts at the end of the year, these counties
were asked: “Were all activities proposed in your OTP application fully implemented as of
6/30/2007, or were some still being developed?”
Twelve of the 32 counties (37.5%) indicated that some activities were still under
development. Therefore performance and outcomes in these counties may not yet reflect the
impact of OTP funds. Still, progress toward two common OTP goals in particular, the
number of clients receiving narcotic replacement therapy (NRT) and the number receiving
residential treatment, can be easily quantified.
In the 13 counties that indicated OTP funds would be used to increase Narcotic Treatment
Program (NTP) slots, unique clients receiving NRT treatment increased from 74
maintenance clients in 2005/2006 to 146 maintenance clients in 2006/2007, representing an
increase of 72 maintenance clients in these counties, or 97.3%.1 Use of NTPs is a promising
practice that will be discussed further in Chapter 8.
In the 33 counties that indicated OTP funds would be used to increase residential treatment
capacity, unique clients receiving residential treatment increased from 4,024 in 2005/2006 to
4,351 in 2006/2007, representing an increase of 327 clients, or an increase of 8.1%.2 More
information on residential treatment can be found in Chapter 9.
Show rates from referral to treatment entry remained nearly the same between 2005/2006
and 2006/2007 (71.3%, 69.3%, respectively) in the 39 OTP counties. However, as discussed
previously (see Chapter 1 and Appendix 1.3) these results must be interpreted with caution
since recent substantial changes in data collection renders comparisons of small show-rate
differences between recent years meaningless. Also, because not all counties had finished
implementing their OTP activities, and not all activities were intended to improve show
rates, show rates may not be the best measure of OTP progress.

Selection of Promising Practices for Additional Study
To generate additional ideas to improve Prop 36 implementation, UCLA examined the
results of interviews conducted with Prop 36 stakeholders as part of a separate study on Prop
36,3 and recommendations from a panel of senior UCLA researchers convened to
recommend the most promising topics based on research literature.4

1

Two small counties reported admissions to CADDS using the same county code and could not be
differentiated. Therefore clients from both counties are included in these counts even though one county
received OTP funding while the other county did not. Together, clients from these two counties account for
less than 2% of NTP client counts over all OTP counties.
2
Two small counties reported admissions to CADDS using the same county code and could not be
differentiated. Therefore clients from both counties are included in these counts even though one county
received OTP funding while the other county did not. Together, clients from these two counties account for
less than 1% of residential client counts over all OTP counties.
3
Treatment System Impact study, Principal Investigator Yih-Ing Hser.
4
The expert panel consisted of Drs. M. Douglas Anglin, Yih-Ing Hser, Christine Grella, Michael Prendergast,
and Richard Rawson.

102

Recommendations from these three sources converged on several areas of particular interest:
continuing care, employment assistance, narcotic replacement therapy, residential programs,
drug testing and sanctions, service integration for the mentally ill, and process improvement.
Given limited evaluation time and resources, UCLA and ADP agreed to focus on several of
these practices, each of which will be discussed in more detail in the following chapters in
this section of the report. The practices are: Employment assistance (Chapter 3), process
improvement (Chapter 4), narcotic replacement therapy (Chapter 5), residential treatment
(Chapter 6), and drug testing and sanctions (Chapter 7). A discussion of service integration
for the mentally ill is also included within discussion of the homeless mentally ill population
(Chapter 4).

Conclusions and Recommendations
Preliminary results indicate success for OTP for the targeted improvements that could be
readily measured. Additional passage of time will allow further evaluation of OTP counties
by allowing a period for counties to fully implement plans and to allow use of additional
measures that require follow-up periods, such as time in treatment and re-arrest rates.
While early results suggest that OTP may be moving Prop 36 implementation in a positive
direction, the pursuit of new promising practices, as well as a better understanding of the
strategies that are already underway must continue. The following chapters provide
preliminary assessments of a number of evidence-based strategies that hold strong potential
for further improving Prop 36 implementation. Some of these practices are already listed as
suggested strategies for which OTP funds may be used (narcotic replacement therapy,
residential treatment, and drug testing and sanctions). At a minimum, UCLA recommends
the addition of employment assistance and process improvement to this list. However, all of
the strategies described will require sustained efforts to maintain. Since, in the past, OTP
funds have been targeted at new or expanded activities, a more permanent source of funding
or a change in OTP goals may be required to sustain these improvements over time.

103

104

Chapter 6: Employment Assistance in Proposition 36
Elizabeth Evans, M.A.
At treatment entry, approximately one-third of Prop 36 offenders are working, a larger
proportion than other non-Prop 36 groups also entering treatment.
Prop 36 offenders who are not working fall into two distinguishable groups.
Unemployed individuals are characterized as being younger and typically assigned to
outpatient care. Individuals who are not in the labor force are older, more likely to be
women, African American, cocaine or heroin users, engaged in daily use, have a longer
drug use history, inject drugs, and are assigned to residential treatment.
During the 5 years of Prop 36 implementation, an increasing number of treatment
programs reported providing employment services, primarily off-site by referral to a nondrug treatment agency. In 2007, about 77% of counties offered employment services to
Prop 36 clients and a fairly wide range of employment-related service types were
reported to be available. However, very few Prop 36 treatment clients reported receiving
employment services (13%) in the three months following treatment assessment and the
amount of services that were received was low (4.8 times). Receipt of employment
services was associated with more severe employment and family problems, assignment
to a residential treatment setting, a greater desire for employment services, and a
race/ethnicity of “other.” Notably, significantly more of the Prop 36 clients who actually
received employment services also completed drug treatment successfully.
Across counties, improvements in employment status from treatment intake to discharge
were small, with many offenders dropping out of the labor force altogether during this
period. Longer term employment outcomes were more promising. At one-year postassessment, about half of offenders were employed or had been paid for work in the prior
month.
The likelihood of being employed one year after Prop 36 treatment assessment was
increased by treatment completion or treatment retention of > 90 days and Hispanic
race/ethnicity, and was decreased by an older age, residing in a particular county, greater
employment problem severity, and receipt of “other” services (including public
assistance).
County stakeholders identified several barriers and promising strategies for addressing
employment needs among the Prop 36 population. While understanding the relative
effectiveness of each approach requires further study, these experiences constitute
potential strategies for future Prop 36 planning and programming.

UCLA’s Prop 36 report for 2004 (Longshore et al., 2005) showed improvements in client
employment status one year post initial assessment, particularly among treatment completers.
However little statewide information on factors associated with improvements in

105

employment status over time has been provided (e.g., offender characteristics, treatment
retention, receipt of employment services). Furthermore, the employment needs, services
utilization, and related outcomes of Prop 36 offenders have been little studied. Evidence
indicates that less than 15% of Prop 36 treatment clients receive employment-related
services, clients who do receive employment services typically receive less than one service
in the first three months of treatment, and that directed program services could be better
matched to meet employment and other needs (Hser et al., 2007b). These findings are of
particular concern since employment is one of the few primary factors associated with
treatment success in the three months following Prop 36 treatment entry (Hser et al., 2007a).
Ample research has demonstrated the strong positive association between employment and
substance abuse treatment outcomes (Buck, 2000; SAMHSA, 2000). Factors contributing to
improved employment outcomes after substance abuse treatment include high motivation for
employment at treatment intake, pre-treatment employment, on-site services, matching
employment services to individual client need, treatment completion or retention of 90 days
or more, and close coordination promoting employment by state and local agencies,
community providers, and employers. Additionally, some studies suggest that criminal
justice-involved populations may be more highly motivated to gain employment and
experience better employment outcomes than other clients, due to external pressures such as
meeting requirements set by probation or parole or to avoid potential incarceration (Magura
et al., 2004). It is clear that vocational and employment training can positively impact client
outcomes and service models frequently include similar elements (e.g., work readiness
education, job seeking skills training, job placement assistance, case management, supported
work) (Hall et al., 1981; Kemp et al., 2004; Kidorf et al., 2004; Staines et al., 2004).
However, there is no generally accepted vocational rehabilitation or employment assistance
model for use with substance abuse treatment clients, particularly those who are also
offenders whose criminal histories present significant employer concerns.
This chapter provides more in-depth information on a range of topics related to employment
among Prop 36 clients. First, information on the employment status of clients at treatment
intake and discharge for each of the six years of Prop 36 implementation is presented. For
context and comparison purposes, data is provided by type of employment status (i.e.,
employed, not employed, and not in the labor force) and also by treatment referral type (Prop
36 probation and parole, non-Prop 36 criminal justice, non-criminal justice). Changes in
employment status from intake to discharge are also shown, as is county variation in those
changes. Second, using the most recent year of available data (fiscal year 2005-2006),
information on the characteristics of Prop 36 offenders by employment status at treatment
intake is presented. Third, the type and amount of employment services provided under Prop
36 are described as reported both by county stakeholders and by Prop 36 clients. Differences
in characteristics between clients who received employment services and those who did not
receive employment services are shown. Also discussed are factors associated with
employment services utilization and positive employment outcomes one-year after treatment
assessment. Fourth, county stakeholder perspectives on successes, barriers, and lessons
learned from addressing Prop 36 employment issues are summarized. The chapter concludes
with a discussion of promising practices and issues for further examination.

106

Sources of data for this section include responses to the Prop 36 Stakeholder Surveys and
Focus Groups (see Appendix A & C respectively), analysis of CADDS and CalOMS data,
and data provided by the Treatment System Impact and Outcomes of Prop 36 (TSI) study
(see Appendix 6.1). In keeping with CADDS definitions (California Department of Alcohol
and Drug Programs, 2001), throughout most of this section, individuals who are working
full-time or part-time are categorized as “employed,” those who are not employed but
actively seeking work are included in “unemployed,” and individuals who are not employed
and not seeking work are coded as “not in the labor force.”

Employment Status of Prop 36 Clients, 2001-2006
Figure 6.1 shows the percentage of clients employed (full- or part-time) at treatment intake,
by the referral source indicated in CADDS/CalOMS. For each year of Prop 36
implementation, approximately two-thirds of Prop 36 offenders were not working at
treatment entry, however more Prop 36 probationers and parolees (approximately one- third)
reported being employed than other types of clients (approximately one-quarter or less)
entering treatment, with more Prop 36 probationers being employed than any other group in
three of the five years examined. A similar pattern was evident at treatment discharge
(Figure 6.2).

Figure 6.1
Percentage of Treatment Clients Employed at Intake
by Referral Source
34

25
.8

9

18
.9

18
.

.7

3

23
.

23
.6
17

20

8

.1
32

29
.

28
.9
19

.2

20
.5

24

.6

26
.5
26
.6

Percent of treatment clients

30

.9
31
.0

31
.3

33
.

5

.4
34
.8

40

0
7/1/01 - 6/30/02
(N = 164,706)

7/1/02 - 6/30/03
(N = 167,340)

Prop. 36 probation

7/1/03 - 6/30/04
(N = 162,488)

Prop. 36 parole

7/1/04 - 6/30/05
(N = 161,334)

Criminal justice non-Prop. 36

107

7/1/05 - 6/30/06
(N = 155,859)

Non-criminal justice

Figure 6.2
Percentage of T reatm ent C lients E m ployed at D ischarge
by R eferral Source
32
.6
31
.3

32

.6

32
.8

.1

.3

26

27

.9

.1

27

.7
18

.0
18

7/1/03 - 6/30/04
(N = 149,923)

7/1/04 - 6/30/05
(N = 142,746)

7/1/05 - 6/30/06
(N = 75,854)

.7

20

17

18

.9

20

.6

27

.1

28

28

.2

.3

31

.2
29
.5

29

Percent of treatment clients

32

.9

34

.9

40

0
7/1/01 - 6/30/02
(N = 153,747)

7/1/02 - 6/30/03
(N = 157 ,196)

P ro p. 3 6 pro ba tio n

P ro p. 3 6 pa role

C riminal jus tic e non-Pro p. 3 6

N on-crimina l jus tic e

Figure 6.3
Change in Employment Status of Treatment Clients
from Intake to Discharge by Referral Source
3.6

3.5
3.0

2.7 2.9
2.0

Percent change

1.8

1.5

1.4

1.5

0.5
0.1

0.1

0.0
-0.3

-0.1

-0.3

-1.8

-1.5

-3.4
7/1/01 – 6/30/02

7/1/02 – 6/30/03

7/1/03 – 6/30/04

7/1/04 – 6/30/05

7/1/05 – 6/30/06

-5
Prop. 36 probation

Prop. 36 parole

Criminal justice non-Prop. 36

Non-criminal justice

Figure 6.4 shows the employment status of the Prop 36 group at intake across years.
Approximately an equal percentage (i.e., about one-third) of Prop 36 offenders were
employed, unemployed, and not in the labor force. More variation in employment status was
evident at discharge (Figure 6.5). Across the years, slightly more Prop 36 offenders at
discharge were not in the labor force (about 38%) than were employed (about 32%), or
unemployed (about 29%).
108

F ig u re 6 .4
E m p loym ent Sta tu s o f P rop . 3 6 C lien ts a t T rea tm en t Intak e
40
35.7

34.9

Percent of treatment clients

32.9 32.2

31.8

31.0

30.9

35.4

34.3 34.7

33.4

32.8

34.4
32.0

33.6

20

0
7 /1 /0 1 - 6/30 /0 2
(N = 2 5,20 8)

7/1/02 - 6 /3 0 /0 3
(N = 36 ,0 28 )

E mplo ye d

7 /1 /03 - 6/30 /0 4
(N = 35 ,9 1 0)

U ne m plo ye d

7/1/04 - 6 /3 0 /0 5
(N = 38 ,8 66 )

7 /1 /0 5 - 6 /3 0/06
(N = 4 0,35 3 )

N o t in la bo r fo rc e

Figure 6.5
Employment Status of Prop. 36 Clients at Treatment Discharge
38.8

40

Percent of treatment clients

34.4

38.9
32.7

38.6
32.1

28.4

38.2

37.7
32.4

32.1
29.3

29.7

29.9

7/1/03 - 6/30/04
(N = 33,537)

7/1/04 - 6/30/05
(N = 35,156)

7/1/05 - 6/30/06
(N = 18,163)

26.8

20

0
7/1/01 - 6/30/02
(N = 23,536)

7/1/02 - 6/30/03
(N = 34,188)

Employed

Unemployed

Not in labor force

Figure 6.6 shows that from 2001 to 2005, the change in employment status from intake to
discharge appeared to be explained mostly by small decreases (3-5%) in the percentage of
unemployed individuals and small increases (3-4%) in the percentage of people not in the
labor force or employed (1-2%). However, beginning in 2005, there was a reduction in the
percentage of Prop 36 offenders becoming employed and unemployed, and an increase in the
percentage of offenders who were not in the labor force. These data indicate that while some
Prop 36 offenders become employed by treatment discharge, more appear to have dropped
out of the labor force altogether.

109

F ig u re 6 .6
C h a n g e in E m p lo y m e n t S ta tu s o f P r o p . 3 6 C lie n ts
fr o m T r ea tm e n t In ta k e to D isc h a r g e
10

Percent change

3 .9

2 .8

1 .8

1 .5

4.1

3 .9

3 .2
1 .1

0 .3

- 2 .0 -2 .1
-3 .1
-5 .0

- 5 .4
7 /1 /0 1 – 6/3 0 /0 2

-5 .0

7 /1 /0 2 – 6/3 0 /0 3

7 /1 /0 3 – 6/3 0 /0 4

7 /1 /0 4 – 6 /3 0 /0 5

7 /1 /0 5 – 6/3 0 /0 6

-1 0
E m plo ye d

U ne m plo ye d

N o t in la bo r fo rc e

There was county variation in changes in percentages of Prop 36 offenders employed from
intake to discharge (Figure 6.7). In some counties, fewer Prop 36 clients were employed at
treatment discharge than at treatment entry while in other counties the percentage employed
remained the same or increased by as much as 20% over the same time period.
F i g u r e 6 .7
C o u n ty V a r ia tio n in C h a n g e in P e r c e n ta g e o f P r o p . 3 6
O ffe n d e r s E m p lo y e d fr o m T r e a t m e n t I n ta k e to D is c h a r g e

Number of counties

40

21

20

20

18

16

15
12

11

10
7

12
9

9

7
4

2

8

7

6

2

1

0
7 / 1 /0 1 - 6 / 3 0 / 0 2
(N = 4 7 )

< 0%

0%

7 / 1 /0 2 - 6 / 3 0 / 0 3
(N = 4 8 )

1 - 5%

7 /1 / 0 3 - 6 / 3 0 / 0 4
(N = 5 0 )

6 - 10%

7 /1 / 0 4 - 6 / 3 0 / 0 5
(N = 5 2 )

11- 20+%

Characteristics of Prop 36 Clients by Employment Status at Treatment Intake
Table 6.1 shows the characteristics of Prop 36 offenders by employment status at treatment
intake. Individuals who were employed full-time or part-time looked similar on most items
analyzed but differed significantly (in part due to large Ns for the sample sizes) for gender;
more part-time workers were women. Unemployed individuals were distinguished primarily
by being slightly younger. The characteristics of individuals not in the labor force were most
distinctive. This group included a greater percentage of individuals who were: older,
women, African American, cocaine or heroin users, daily drug users, drug users for more
years, injection drug users, and those in residential treatment.
110

Table 6.1: Prop 36 Client Characteristics by Employment Status at Treatment
Intake 7/01/05 - 6/30/06 (N = 40,353)
Employed Employed
Not in
Unemployed
Full-time Part-time
labor force
(N=12,910)
(N=9,446) (N=4,442)
(N=13,555)
34.8
35.1
33.5
36.2
Age, mean
13.5
25.1
27.8
36.0
Female, %
Race/ethnicity, %
44.5
47.7
42.7
43.3
White
8.1
10.4
12.6
19.3
African American
41.1
35.0
37.5
31.0
Hispanic
2.8
3.4
3.5
2.2
Asian
1.0
1.4
1.4
2.0
Native American
2.2
2.3
2.1
1.9
Other
11.3
11.3
11.1
11.1
Education, mean
Referral source, %
86.3
87.3
87.1
86.5
Probation
13.6
12.6
12.8
13.4
Parole
Primary drug, %
60.2
58.4
60.4
51.1
Methamphetamine
11.3
12.0
10.7
16.9
Cocaine
5.1
6.2
7.5
11.0
Heroin
13.3
14.2
13.0
10.8
Marijuana
8.8
7.9
7.1
8.7
Alcohol
1.1
1.0
1.0
1.2
Other
Frequency of primary drug use, %
44.3
42.0
41.6
37.5
None
24.2
24.6
20.5
20.0
1-3 times/month
14.0
13.9
13.0
13.0
1-2 times/week
7.1
8.3
9.7
10.5
3-6 times/week
10.2
11.0
14.9
18.9
Daily
Years since first primary drug use,
13.4
14.0
13.2
15.6
mean
7.3
8.7
10.9
15.9
Injects drugs, %
Modality, %
95.7
96.0
90.0
66.8
Outpatient
2.9
2.4
7.4
27.2
Residential
<1.0
1.0
1.0
1.2
Narcotic replacement therapy
47.7
49.5
51.0
49.6
Prior treatment, %
Source: CADDS

111

Employment Services Provided to Prop 36 Clients
A UCLA ISAP study called Treatment System Impact and Outcomes of Prop 36 (TSI)
collected information from treatment programs on the types of services available to Prop 36
clients (for more information about the TSI study, see Appendix 6.1). Analysis of services
provided in the year 2000 (i.e., the year before Prop 36 implementation) compared to
services provided in 2001, 2002, and 2005 showed that an increasing number of treatment
programs, approximately two-thirds in 2005, reported that they provided services to address
employment problems.
Of services that could potentially impact employment outcomes, some were primarily
provided on-site (e.g. ., transportation) but most others (i.e., literacy training, GED
education, employment assistance/vocational training) were provided off-site, by referral to a
non-drug treatment agency. In 2005, literacy training and GED education were provided onsite by 10% of programs and by referral by 67% of programs. Transportation assistance was
provided on site by 51% of programs and by referral by 12% of programs. Employment
assistance and vocational training were provided on-site by 26% of programs and by referral
by 46% of programs. However, follow-up data on whether clients actually used the referred
services was not known to program staff completing the survey.
Prop 36 stakeholder focus group participants discussed the types of employment services that
were available to some Prop 36 clients in 2006-2007. Services included:
•
•
•
•
•
•
•
•
•
•
•

•
•

Assessment of need for vocational services
GED education
Access to computers and job listings
Vocational education groups
Employment workshops
Appointment with a full-time on-site professional job counselor
Job fairs
Bus passes to travel to trainings and job interviews
Assistance with physical appearance and presentation
Lessons on navigating the SSI and public aid system
Referral to services provided by other agencies (e.g., Employment
Development Department, Department of Vocational Rehabilitation,
CalWORKS)
Specialized services for parolees
Linkage to agencies that provide voicemail services

As a complementary source of information, county lead agency stakeholders, court
administrators, and treatment programs who responded to the 2007 Prop 36 Survey indicated
that employment services were available to Prop 36 clients in 77% of counties (see Table
6.2). Both county lead agency representatives and treatment providers indicated that
employment services were mostly provided off-site at a non-drug treatment agency, and
included a range of service types. Court administrators indicated that about half of courts
112

assigned Prop 36 clients to employment services if needed, while about one-third of courts
assigned clients to receive literacy services. It must be noted, however, that no information
was available on the number of clients who actually utilized and/or benefited from
employment services. Furthermore, across respondent groups, the employment-related
services provided most infrequently included job placement, literacy training, job skills
training, and GED education. Also, when treatment provider respondents were asked to
identify the top three services of most urgent or pronounced need (data not shown), few (i.e.,
less than 5%) identified employment-related services, however of those that did rank
employment services as an area of need, job placement and job skills training were identified
most frequently as urgent.

Table 6.2: Employment services available to Prop 36 offenders, %
7/1/06-6/30/07
County
Treatment
Court
lead agency providers
(N = 27)
(N = 48)
(N = 86)
77.1
--55.6
Employment services are provided
Location of employment services
On-site at drug treatment program

40.5

---

---

Off-site at non-drug treatment program

81.1

---

---

Job readiness assessment

78.4

81.4

---

GED education

75.7

69.8

---

Vocational counseling

81.1

81.4

---

Job-seeking skills training

89.2

80.2

---

Resume assistance

83.8

83.7

---

Job skills training

64.9

73.3

---

Information on job openings

91.9

80.2

---

Job placement

48.6

62.8

---

Literacy training

---

64.0

33.0

Other

27.0

---

---

Type of employment services

Source: Prop 36 stakeholder surveys
Employment Services Utilization
Using TSI client data, UCLA examined the characteristics of Prop 36 clients who reported
receiving employment services (N =192) and compared them to clients who said they did not
receive employment services (N = 1,261) (Table 6.3). Receipt of employment services was
defined as having seen someone (e.g., employment specialist, counselor, or social worker)
regarding employment opportunities, training, or education in the three months following the
Prop 36 assessment for treatment. Of the total sample, more than one-third was employed

113

Table 6.3: Characteristics of Prop 36 Treatment Assessment of Clients
Who Did and Did Not Receive Employment Services
Received employment services?
Yes (N=192,
No (N=1,261
Total
13%)
87%)
(N=1,588)
Age, Mean (Standard Deviation) [M(SD)]
37.6 (9.7)
36.7 (9.7)
36.8 (9.8)
Race,%
White
45.3
51.4
50.6
Hispanic
25.0
24.8
24.8
Black
19.2
18.0
18.1
Other
10.4
5.7
6.3
Women, %
34.3
28.3
29.1
Education, M (SD)
11.7 (1.6)
11.7 (1.0)
11.7 (1.9)
Married, %
13.3
15.1
14.9
Homeless, %
11.7
8.1
8.6
Arrested in past 30 days, %
20.8
23.9
23.5
Times arrested in lifetime, M (SD)
9.6 (13.0)
8.8 (11.7)
8.8 (11.9)
Months incarcerated in lifetime, M (SD)
26.3 (32.7)
25.3 (34.1)
25.3 (33.7)
County, %
County 1
22.9
26.0
25.6
County 2
20.8
22.2
22.0
County 3
23.9
24.2
24.2
County 4
9.9
11.1
10.9
County 5
22.4
16.3
17.1
Drug use and treatment
Addiction Severity Index, M (SD)
Alcohol
0.10 (0.19)
0.10 (0.17)
0.10 (0.17)
Drug
0.14 (0.11)
0.12 (0.11)
0.13 (0.11)
Employment**
0.76 (0.25)
0.70 (0.29)
0.71 (0.28)
Family**
0.19 (0.21)
0.15 (0.19)
0.15 (0.19)
Legal
0.26 (0.18)
0.26 (0.18)
0.26 (0.18)
Medical
0.25 (0.32)
0.24 (0.33)
0.23 (0.32)
Psych
0.19 (0.22)
0.16 (0.21)
0.16 (0.21)
Primary drug, %
Methamphetamine
47.8
51.9
51.4
Cocaine
14.8
11.5
12.0
Marijuana
12.7
12.1
12.2
Alcohol
6.3
7.9
7.7
Heroin
11.7
8.4
8.8
Other
3.1
2.7
2.8
Used primary drug in past 30 days, %
52.0
49.2
49.5
Modality, %**
Narcotic replacement therapy
6.4
4.1
4.4
Outpatient
66.4
78.3
76.8
Residential
27.0
17.5
18.7
Number of prior treatments, M(SD)
2.9 (5.1)
2.2 (3.8)
2.3 (4.0)
(p=0.06)

114

full- or part-time (38.6%), one-third was not in the labor force (32.5%), and more than onequarter (28.7%) were unemployed (i.e., looking for work).
Despite the relatively high reports by lead agencies and programs, very few Prop 36 clients
reported receiving employment services (13%). On most indicators, the characteristics of
clients who received employment services were very similar to the characteristics of those
who did not receive employment services. Clients who received employment services did
have more severe employment (ASI composite score of 0.76 vs. 0.70) and family (ASI
Composite score of 0.19 vs. 0.15) problems and more of those treated in a residential as
opposed to an outpatient setting received employment services (27.0% vs. 17.5%).
More differences were revealed by analysis of employment-related variables (Table 6.4).
Compared to clients who did not receive services, fewer clients who did receive services
were employed at assessment (29.2% vs. 40.1%), a smaller percentage had been paid for
working in the prior 30 days (24.3% vs. 37.8%), and clients who received employment
services had fewer days of paid work (3.5 vs. 5.7) in the prior 30 days. More of the clients
who received employment services also received income from welfare (15.8% vs. 8.7%), and
fewer received income from a pension (6.8% vs. 13.2%) or employment (26.4% vs. 39.0%).
Furthermore, fewer of the clients who received employment services reported having another
person dependent on them for support (22.3% vs. 29.7%) but more of them indicated that
they wanted employment services (59.4% vs. 45.0%).

Table 6.4: Employment Status Income Sources
Received employment services?
Yes
No
Total
(N=192) (N=1,261) (N=1,588)
Current employment status, %**
Employed
29.2
40.1
38.6
Unemployed
40.9
26.9
28.7
Not in labor force
29.7
32.8
32.5
Paid for work in past 30 days,%**
24.3
37.8
36.0
Days paid for working in past 30 days, M (SD)**
3.5 (7.4) 5.7 (9.0)
5.5 (9.0)
$466
$590
$586
Income in past 30 days, Mean (SD)
(1588.6) (943.4)
(1238.1)
Income source, %
Employment**
26.4
39.0
37.3
Unemployment
2.6
2.4
2.4
Welfare**
15.8
8.7
9.7
Pensions, SSI**
6.8
13.2
12.3
Family, friends
28.5
27.9
28.0
Someone contributes to support,%
50.5
46.4
46.9
Other people depend on person for support, %*
22.3
29.7
28.7
Receives psychiatric pension, %
4.4
5.9
5.7
Had employment problems in past 30 days, %
45.1
38.0
38.9
Wants employment services, %**
59.4
45.0
46.8
*p<0.05; **<0.01, Source: TSI

115

On average, Prop 36 clients received services a mean of 4.8 times for employment problems
over the three months following assessment for treatment (Table 6.5). Compared to clients
who did not receive employment services, more clients who received employment services
also saw a professional regarding unemployment benefits (30.2% vs. 5.9%), and more had
individual or group sessions to discuss employment and support problems (59.3% vs. 9.8%).

Table 6.5: Receipt of Employment and Other Related Services Over 3 Months
Following Prop 36 Treatment Assessment
Received employment services?
Yes
No
Total
(N=192)
(N=1,261)
(N=1,588)
Employment & related services
Number of times received employment
4.8 (9.4)
0.0
0.6 (3.7)
services, Mean (SD)**
Saw unemployment specialist, counselor,
30.2
5.9
9.1
social worker, %**
Had individual or group session about
59.3
9.8
16.3
employment/support problem, %**
Been in school or training, %**
7.8
3.9
4.4
Other services, %
Medi-Cal
16.1
13.9
14.2
General relief**
11.4
6.0
6.7
Food stamps**
15.1
8.2
9.1
Public assistance**
7.8
2.7
3.3
Women, Infants, and Children (WIC)
2.9
1.3
1.5
Employment Development Dept (EDD)**
5.2
1.9
2.3
Supplemental Security Income (SSI)**
4.1
10.0
9.3
Child Protective Services (CPS)
1.1
1.1
1.1
Other services
2.0
1.7
1.8
Assistance with, %
Housing
7.3
4.7
5.0
Transportation**
17.4
6.5
7.9
Other basic needs**
10.5
5.1
5.8
*p<0.05; **<0.01, Source: TSI
There were also some differences in the type of additional employment/support-related
services received. Individuals who obtained employment services also received more
services related to General Relief (11.4% vs. 6.0%), food stamps (15.1% vs. 8.2%), public
assistance (7.8% vs. 2.7%), EDD services (5.2% vs. 1.9%), transportation (17.4% vs. 6.5%),
and other basic needs (10.5% vs. 5.1%). Conversely, more clients who did not receive
employment services got SSI services (10.0% vs. 6.1%).
Further analysis of TSI data revealed a few significant factors associated with receipt of
employment services (Table 6.6). Clients with a race/ethnicity of “other” and those who
reported wanting employment services at intake were more likely to receive employment

116

services. Clients assigned to outpatient as opposed to residential treatment were less likely
to receive employment services.

Table 6.6: Multivariate Analysis Predicting Receipt of Employment Services
(N = 1,350)
Odds Ratios1
Age
1.00
African American (vs. White)
0.90
Hispanic (vs. White)
1.24
Other (vs. White)*
1.91
County 1 (vs. County 5)
1.08
County 2 (vs. County 5)
1.37
County 3 (vs. County 5)
1.14
County 4 (vs. County 5)
1.45
Female (vs. Male)
1.29
Outpatient (vs. Residential)**
0.85
ASI Employment Composite Score
1.07
Paid for work in 30 days prior to intake
0.68
Wants employment services**
1.64
*p<0.05; **<0.01
Source: TSI
Employment Outcomes
Next UCLA examined differences in outcomes between those receiving employment versus
those who did not at 12 months after Prop 36 assessment (Table 6.7). At the 12-month
follow-up, the two groups demonstrated similar improvements in most areas examined.
About half of offenders were employed and had been paid for work in the prior month, 10%
or less had been arrested, and 15% had used their primary drug during the past 30 days.
Most notably, compared to their counterparts, significantly more of the clients who received
employment services also completed drug treatment. Also, although not statistically
significant, fewer of these clients were arrested, more had stayed in treatment for at least 90
days, they had spent more days in treatment, and more had completed the Prop 36 program.
Examination of the degree of change from assessment to one year later indicates that clients
who received employment services experienced greater improvements in some areas. For
example, more clients who received employment services had become employed at followup compared to baseline (28.0% increase) than those who did not receive employment
services (16.0% increase). Similarly, greater change occurred from baseline to follow-up
1

The odds ratio is a way of comparing whether the probability of a certain event is the same for two
groups. An odds ratio of 1 implies that the event is equally likely in both groups. An odds ratio greater
than one implies that the event is more likely in the first group. An odds ratio less than one implies that the
event is less likely in the first group.

117

among people who obtained employment services, compared to people who did not receive
employment services, when examining the change in percentages of people paid for work

Table 6.7: Client status 12 months after Prop 36 treatment assessment
Received employment services?
Yes
No
Total
(N=192)
(N=1,261)
(N=1,588)
Employment status, %**
Employed (full/part-time)

57.2

56.1

56.2

Unemployed

11.5

8.8

9.1

Not in labor force

31.2

35.0

34.5

Arrested, %

6.9

10.3

9.8

Used primary drug, %

16.0

14.6

14.8

Paid for work, %

57.4

54.8

55.2

8.3 (8.5)

9.3 (9.6)

9.1 (9.5)

$1,065.0
(1,003.8)

$1,041.9
(988.2)

Days paid for work, Mean (SD)
Income, Mean (SD)

$922.1 (889.0)

Income source, %
Employment

56.9

54.6

54.9

Unemployment

<1.0

1.7

1.5

Welfare

13.7

9.3

9.9

Pensions, SSI

13.2

13.8

13.7

Family, friends**

55.1

41.3

43.2

Alcohol

0.02 (0.07)

0.03 (0.08)

0.03 (0.08)

Drug

0.03 (0.07)

0.02 (0.06)

0.03 (0.06)

Employment

0.59 (0.31)

0.58 (0.33)

0.58 (0.33)

Family*

0.07 (0.11)

0.06 (0.11)

0.06 (0.11)

Legal

0.09 (0.16)

0.10 (0.16)

0.10 (0.16)

Medical

0.12 (0.24)

0.12 (0.24)

0.12 (0.24)

0.13 (0.18)
60.5

0.12 (0.18)
55.0

0.12 (0.18)
55.7

154.2 (123.9)

134.4 (116.2)

134.1 (115.3)

51.3
42.7

38.5
37.1

40.2
37.9

Addiction Severity Index, Mean (SD)

Psych
Treatment retention >90 days, %
Days in treatment, Mean (SD)
(p=0.06)
Completed drug treatment, %**
Completed Prop 36 program, %
*p<0.05; **<0.01, Source: TSI

118

(33.1% increase vs. 17.0%) and receiving income from employment (30.5% vs. 15.6%),
pension/SSI (6.4% vs. <1%), and family/friends (26.6% vs. 13.4%). Notably, the ASI
Composite Score for problems in the employment domain decreased by 0.17 for clients who
received employment services, greater than the decrease of 0.12 among people who did not
receive services.
Predictors of Employment Outcomes
Analysis of factors associated with being employed 12 months after treatment assessment
showed several significant effects (see Table 6.8). Specifically, the likelihood of being
employed one year after Prop 36 treatment assessment was increased by treatment
completion or by retention of > 90 days and Hispanic race/ethnicity, and decreased by older
age, residing in County 1, a higher ASI Employment Composite Score indicating greater
severity, and receipt of “other” services and public assistance (i.e., Medi-Cal, general relief,
food stamps, public assistance, etc., and also assistance with housing, transportation, and
other basic needs).

Table 6.8: Multivariate Analysis Predicting Employment 12 Months After Intake
(N = 980)
Odds Ratios
Age

0.98**

Hispanic (vs. White)

1.65**

Other (vs. White)

0.83

County 1 (vs. County 5)

0.44**

County 2 (vs. County 5)

0.78

County 3 (vs. County 5)

0.68

County 4 (vs. County 5)

0.80

Female (vs. Male)

0.81

Outpatient (vs. Residential)

0.98

ASI Employment Composite Score

0.45**

Want employment services

1.11

Received employment services

1.08

Received “other” services

0.57**

Completed treatment or retention > 90 days

1.80**

*p<0.05; **<0.01, Source: TSI
The greater likelihood of positive employment outcomes among Hispanic groups may be
explained, in part, due to better access and motivation for work because of cultural and
family obligations, and Hispanics may also exhibit a greater willingness than other
racial/ethnic groups to perform unskilled work. Additional information is needed to better
understand contextual factors that predict employment.

119

Barriers and Facilitators to the Provision of Employment Services under Prop 36
Barriers
Prop 36 stakeholder focus group participants attributed employment outcomes to several
county-level implementation and operation factors. Employment barriers clustered into two
broad categories. Client-centered barriers included the client’s limited ability to secure
employment for reasons such as: fear of reporting a felony conviction on job applications,
recurring relapse to substance abuse, spotty or no employment experience, co-occurring
metal health disorders, feelings of hopelessness and inability to change employment
prospects, and difficulty balancing the obligations of treatment, employment or vocational
training, and a personal life.
System-centered barriers describe realities that made it difficult for employment services to
be provided by treatment or utilized by clients. Examples include: client work schedules that
conflict with court appearances or treatment requirements, treatment programs’ inability to
bill for employment services, lack of transportation for clients, limited resources and funding
among treatment programs to address employment needs, and the tradition among
professionals in the treatment field of viewing employment services as ancillary, and thus of
lower priority, rather than primary care.
Some stakeholders commented on the generally high unemployment rate among the general
population in their county, implying that this atmosphere made it especially difficult for Prop
36 clients to obtain work. As shown in Appendix 6.2, there is some variation in the
unemployment rate among the general population by county. In 2006, unemployment rates
ranged from 3.4% to 15.3%, with 22 counties having an unemployment rate below the
statewide rate of 4.9%, and the remaining 36 counties reporting an unemployment rate above
the statewide rate. In addition, in some counties, especially small or rural counties, there are
few job prospects and the jobs that are available receive many applicants, making getting a
job a real challenge for which most Prop 36 clients are not prepared to undertake.
Others emphasized that the Prop 36 program itself can make it very difficult to secure or
maintain a job, given the many requirements Prop 36 places on clients’ time given its legal
and treatment provisions. Prop 36 treatment usually creates a highly structured day to keep
clients engaged in the recovery process. Clients generally are not free to make a work
commitment until after remaining in Prop 36 for some time, up to nine months or more.
Some counties do provide services in the evenings or during other times when clients are
likely to not be working, however many counties, especially small and rural counties, simply
do not have the resources to offer a wide spectrum of services at different times of the day.
Several focus group participants indicated that drug treatment professionals are not trained to
provide employment or vocational services and so counselors may lack needed expertise to
effectively address employment problems. Moreover, the primary goal of most treatment
counselors is to stay focused on facilitating treatment and recovery. Other focus group
participants added that exploration of employment options is a natural part of recovery,
meaning the more clients are integrated into treatment and are “solid in recovery” the more
likely they are to become educated about employment options and other opportunities. Some
county stakeholders felt more services to address employment problems are provided than

120

what is contractually required and the county is not compensated for that “extra work.” In
addition, while most agreed that employment contributes to greater client self-esteem, it must
be remembered that paychecks can be a trigger to drug purchase, use, and relapse.
Finally, changes in Prop 36 funding from year to year has resulted in an expansion and then
contraction of employment services resources in some counties. Fluctuations in funding and
associated programming have meant that some counties created employment-focused
strategies that were then dismantled shortly thereafter. These changes also often occurred
just as stakeholders perceived that the strategies that had been implemented were taking real
effect. Stability of funding for sustaining all pertinent recovery processes was repeatedly
emphasized.
Focus groups participants also explained that employment outcomes are typically tracked via
existing CADDS and CalOMS data systems. One county noted that for 2006, 25% of Prop
36 clients who entered treatment had a job at intake and the employment rate increased to
41% at discharge. Notably, however, employment information may be missing because a
CADDS/CalOMS discharge record has not been submitted. Also, CalOMS is a relatively
new data system and stakeholders noted that some treatment providers may not yet know
how to data enter employment measures into the system. Other stakeholders noted that
employment status is often self-reported by clients and that for many clients, especially those
in outpatient treatment, having a job and not having a job are fluid concepts, especially for
jobs involving unreported income. In other counties, probation keeps monthly employment
statistics with a data system that is specific to Prop 36 offenders. Many stakeholders agreed
that employment status at treatment discharge is usually not the best indicator of
employment outcomes simply because most clients have only just achieved stable recovery
and have not yet had time to become employed. Stakeholders suggested that measuring
employment outcomes over a longer time period, for example at least nine months after
treatment entry, would probably result in more accurate information on changes to clients’
employment status following Prop 36 participation.
Facilitators
Stakeholder focus group participants also identified some Prop 36 program elements that
facilitated improving client employment outcomes. For example, several participants said
that probation officers routinely supervise employment status and encourage clients to secure
a job or to improve their employment situation. Stakeholders agreed that it was often useful
to draw upon employment resources routinely made available by criminal justice agencies
like probation and parole offices as well as local Sheriff’s departments.
In another example, stakeholders found it useful to require clients to seek employment as a
part of their treatment plan or to be employed in order to fulfill treatment completion
requirements. Stakeholders felt that making job-seeking a part of treatment requirements
seemed to elevate the importance of employment in the minds of clients while also
ameliorating client complaints about having to balance multiple demands on their time.
Also, to facilitate employment seeking, some treatment centers hosted networking gatherings
that included former treatment clients who have secured employment, so that current clients
could shift into a different social network, i.e., a network that values work over using drugs.

121

To make the most of scarce resources, stakeholders reported targeting particular clients for
employment services. Some counties screened clients for employment alternatives based on
need for SSI or other similar assistance, current placement on SSI, and expressed interest in
obtaining a job or a better job. Others noted the importance of having a designated staff
person to perform social work activities to ensure clients were linked with publicly supported
services and programs (e.g., SSI or WIC) based on eligibility.
Other successes were reported primarily related to actively removing barriers to
employment. For example, to aid client employment efforts, some treatment programs
provided:
•
•
•
•
•
•
•
•
•
•

GED graduation ceremonies
regularly scheduled vocational education sessions preferably weekly and
available in the evenings to allow for clients who are already employed to
attend
job lists of “felon-friendly” employers or seasonal employers who may be
more willing to hire individuals with a criminal history
references to employers who had a personal relationship with treatment staff
(i.e., a friend or relative)
on-site, comprehensive, and integrated employment services (i.e., “one-stop
shopping”, provision of a wide array of services)
appointments with a full-time on-site professional job counselor
counseling to overcome client fear of reporting their criminal history on job
applications
access networks of alumni (including 12-step groups) who provide job search
assistance and contacts in existing job networks
hosting “social events” for clients to make contacts with employed peers
linkage to local community colleges to pursue educational interests

Focus group participants generally agreed that drug treatment cannot simply be about getting
clients to not use drugs. Several treatment professionals said that changing substance abuse
and criminal behavior is possible, especially if those changes translate into positive rewards
and an improved quality of life as evidenced by enhanced self esteem, career, employment,
ability to pay bills, a new social network, access to healthcare, and connecting to society.
Employment is a key factor in giving clients an alternative identity to “drug user.”
Furthermore, treatment providers emphasized that providing employment services means not
just helping clients who don’t have a job find employment, but also helping underemployed
clients meet their potential. In particular, treatment programs can help clients to understand
that “all is not lost” with a felony record, a valuable insight for motivating active change.
County stakeholders felt that Prop 36 clients have been receptive to employment services
and structured vocational coaching assistance, viewing access to such services as an
advantage. Focus group participants strongly felt that on-site provision of comprehensive
and integrated employment services increases the likelihood that clients will utilize such
services and remain engaged in the Prop 36 program. In some counties, probation has
played a key role in providing employment opportunities to Prop 36 clients through
vocational rehabilitation connections and encouragement for remaining employed. Also, it is
122

important that Prop 36 services (e.g., trainings, education, counseling sessions) and
requirements (e.g., drug testing, court appearances) accommodate the schedule of clients
who are working or attending school. Some county stakeholders also indicated that fees for
urine testing increase the stress felt by clients, especially if clients are unemployed and nonpayment of fees means the client cannot graduate.

Promising Practices and Issues for Further Examination
The data presented above illustrate the many complexities of addressing, and assessing,
employment needs, services utilization, and outcomes among Prop 36 clients. However,
analyses featured in this chapter revealed several practices that appear worthy of further
examination for potential replication. Promising practices for priority assessment include:
•

•
•
•

•
•
•

•
•

At treatment entry, assess clients for need for employment services. The
assessment should consider not just the individual’s current employment
status, but also their marketable skills, recent work history, and desire for
employment services.
Recognize that the employment needs of individuals who are unemployed
may be different from those of individuals who are not in the labor force.
Address client fears about disclosing their criminal history to prospective
employers as well as insecurities related to unstable or weak work histories.
Target employment resources to maximize the matching of services to need.
Consider implementing strategies to earmark selected individuals for
employment services, for example based on their current employment status,
recent work history, and desire for employment services.
Provide a broad range of skills training and employment services. Services
should be available on-site at the same location as treatment, or clients should
be transported to and from the location where such services are offered.
Make Prop 36 program requirements flexible enough in access and timing to
accommodate the schedule of clients who are employed.
Consider making employment a criterion for treatment completion and/or
Prop 36 program completion. Ensuring client attendance at employment
enhancement services or extending Prop 36-paid treatment services until
these criteria are met are among the possibilities.
Optimizing utilization of employment resources available through the
criminal justice system, in addition to contacts already used by treatment
agencies.
Measure employment outcomes beyond treatment discharge (e.g., through
post treatment follow-up or via state Employment Development Department
records).

Additional evaluative information is needed on the provision and utilization of employment
services under Prop 36, how employment is impacted by particular factors (such as case
management, client motivation level, on-site services, matching of services to need, and
treatment completion), and whether outcomes are associated with an identifiable collection
of strategies applicable to the Prop 36 population. Future efforts should focus on evaluating
the promising practices identified in this chapter, to identify those that are most effective for

123

improving employment outcomes among the Prop 36 population, and to methods of
successfully transferring such practices into program services.

References
Buck, M. (2000). Getting Back to Work; Employment Programs for Ex-Offenders, Field
Report Series. Philadelphia, PA: Public/Private Ventures.
http://www.ppv.org/ppv/publications/assets/94_publication.pdf
California Department of Alcohol and Drug Programs. (2001). California Alcohol and
Drug Data System (CADDS): Instruction Manual.
Hall et al. (1981). Increasing employment in ex-heroin addicts II: methadone
maintenance sample. Behavior Therapy, 12, 453-460.
Hser, Y.I., Evans, E., Teruya, C., Huang, D., & Anglin, M.D. (2007a). Predictors of
short-term treatment outcomes among Proposition 36 clients. Evaluation and
Program Planning, 30, 187-196.
Hser, Y.-I., Teruya, C., Brown, A.H., Huang, D., Evans, E., & Anglin, E. (2007b).
Impact of California’s Proposition 36 on the drug treatment system: Treatment
capacity and displacement. American Journal of Public Health, 97, 104-109.
Kemp et al. (2004). Developing employment services for criminal justice clients
enrolled in drug treatment programs. Substance Use and Misuse, 39.
Kidorf et al. (2004). Combining stepped care approaches with behavioral reinforcement
to motivate employment in opioid-dependent outpatients. Substance Use and
Misuse, 39.
Longshore, D., Urada, D., Evans, E., Hser, Y.-I., Prendergast, M., & Hawken, A., (2004).
Evaluation of the Substance Abuse and Crime Prevention Act: 2004 report.
Sacramento, CA: Department of Alcohol and Drug Programs, California Health and
Human Services Agency.
Magura et al. (2004) The effectiveness of vocational services for substance users in
treatment. Substance Use and Misuse, 39, 2165-2213.
SAMHSA (2000). Treatment Improvement Protocol (TIP) Series 38: Integrating
Substance Abuse Treatment and Vocational Services. Available at
http://www.ncbi.nlm.nih.gov/books/bv.fcgi?rid=hstat5.chapter.68228.
Staines et al. (2004). Efficacy of the Customized Employment Supports (CES) model of
vocational rehabilitation for unemployed methadone patients: preliminary results.
Substance Use and Misuse, 39.

124

Chapter 7: Treatment Process Improvement Methods and their
Application to Proposition 36
Beth Rutkowski, M.P.H. and Darren Urada, Ph.D.
Process improvement methods pioneered in business and industry settings to increase
efficiency and productivity can be applied to community-based substance abuse treatment
organizations at relatively low cost. The nationwide Network for the Improvement of
Addiction Treatment (NIATx) model of process improvement provides a structure under
which program staff identify needs, determine measurable goals, select and implement
changes, monitor subsequent results, and adjust as necessary. The NIATx model has been
applied repeatedly to achieve goals that would be desirable in Prop 36, such as reductions in
no-show rates and increases in client continuation in treatment.
In 2005-2006 six outpatient treatment programs and one residential program in Los Angeles
County participated in a demonstration project to determine if the model could improve
treatment retention and completion rates. Three of these treatment providers served Prop 36
clients.
The programs used a variety of innovative strategies selected by their own staffs, including
same day assessments, increased contact with prospective clients, consolidated intake
paperwork, incentives, appointment cards, and satisfaction surveys. Most programs were
able to demonstrate modest to marked improvements in no-show rates, counseling session
attendance rates, and continuation rates. Aggregate data from the six outpatient programs
revealed an 80% reduction in assessment no-shows and a 6% increase in 30-day continuation
rates.
In 10-month follow-up interviews, treatment programs reported generally maintaining the
process changes they had implemented during the pilot project. However, continuation of
the process improvement model into new areas was mixed. Maintenance of sustained
improvement efforts will require a permanent infrastructure to support program staff,
especially in the identification and adoption of additional process improvement strategies.
A supportive and controlled roll-out of process improvement techniques throughout a
regional treatment system would be ideal. Participants in the pilot program reported that
guidance from the coordinating Project Director and Process Improvement Coach were
instrumental in their success, and that technical assistance with data collection was key.
Without sufficient levels of such support, new participating programs may not experience the
same results seen in the pilot project.
Process improvement is an evidence-based framework that provides a systematic problemsolving approach that organizations can use to develop a deeper understanding of client
needs, restructure the workflow to more effectively respond to client and staff needs, and
make the most efficient use of available resources (Langley, Nolan, Nolan, Norman, &
Provost, 1996).

125

In Prop 36 there is particular interest in improving processes as they relate to client show
rates at assessment and first treatment contact, engagement and retention in treatment, and
issues contributing to treatment drop-out at any point. UCLA’s survey of Prop 36 treatment
programs (see Appendix B for methods) included questions intended to gauge how providers
were currently assessing process changes in these areas. Of the 86 randomly sampled
programs that returned surveys, 69 (80.2%) indicated that they had implemented changes in
the last year intended to improve show rates, reduce treatment drop out, and/or increase
retention. Nine programs indicated that they had not, while nine did not respond to the
question. Of the 69 programs that indicated making improvements, a large majority reported
judging the impact based on management observation and/or discussions at staff meetings.
Approximately half of the programs reported actually collecting data before and after the
change to measure the effect, which would be a part of any formal process improvement
program (see Table 7.1).
Table 7.1: Responses to “How is the impact of the change typically
assessed?” (n=69)
Top Three Responses

% Yes

Changes are discussed at staff meetings

97.1

Director/management judges the impact based on
observation.

85.5

Outcome data are systematically collected before and after
the change to measure the effect

56.5

Two programs reported using client satisfaction surveys to assess the impact of the change.
One responded “county contract changes.” One indicated using a “multi-dimensional
CQI/performance improvement program.” Continuous Quality Improvement (CQI) is a term
applied to a process improvement approach widely adopted in business before being
introduced to health care by Berwick (1989) and Laffel and Blumenthal (1989). More
recently similar process improvement methods have been successfully applied in drug
treatment programs. The most extensive of these efforts is described below.

The Network for the Improvement of Addiction Treatment (NIATx)
A systematic examination of process improvement strategies and their effects in drug
treatment was initiated nationally in 2003. The Network for the Improvement of Addiction
Treatment (NIATx), a partnership between the Robert Wood Johnson Foundation’s Paths to
Recovery program and the Center for Substance Abuse Treatment’s Strengthening Treatment
Access and Retention (STAR) program, was formed to promote process improvement
specifically in substance abuse treatment.1 The staff of NIATx member programs learn to
apply process improvement principles to improve client engagement and retention in
addiction treatment, while integrating process improvement into program culture. Providers
1

For more information on NIATx, visit www.niatx.net.

126

use process improvement methods to improve services and client attendance by focusing on
four service delivery aims: (1) reducing waiting times from first contact to admission and
receipt of first service; (2) reducing no-shows to assessment or admission interviews; (3)
increasing admissions to the level of funded capacity; and (4) increasing client continuation
rates. NIATx provides a set of tools designed to help treatment providers improve their
programs in these areas and to attain better client outcomes.
The change processes used to achieve these goals include:
•

•
•

•

•

Identify process barriers. This may be done by conducting client interviews
or a program walk-through from a client’s perspective to understand how
current processes facilitate or inhibit treatment goals.
Identify what is to be accomplished and define a reasonable and measurable
goal.
Establish a Change Team to select, adapt if needed, and test the potential
changes identified for addressing targeting problems. The Change Team is
formed by the Executive Director of the organization and a staff member
designated as Change Leader. Effective Change Teams often include a client
or “customer.”
Use a rapid Plan-Do-Study-Act (PDSA) cycle including the collection of data
before, during, and after a change to evaluate whether it resulted in
improvement.
Make adjustments to continuously improve and sustain changes.

The specific processes to be improved are identified and changed based not only on the
identified needs of individual sites, but also on the target goals in each identified area. The
key innovation is the use of a tested model (e.g., the PDSA cycle) to guide process
improvement.
NIATx has produced a series of “aims” and “paths” primers (known as Promising Practices),
which are based on national findings (see the “Promising Practices and Strategies” of the
NIATx website for more information)2. The main conclusion found amongst the
participating NIATx sites is that retention seems to be contingent on having a system in
place that “helps clients develop connections with other people and to a
community…involves developing a sense of inclusion, affinity, belonging, and bonding with
a peer group.” Change Leaders were encouraged to adopt continuation strategies that fit with
their program’s particular situation, not necessarily a strategy that has already been tested in
NIATx.
To gauge current awareness and implementation of NIATx, which distributes its material for
free from its website, among Prop 36 substance abuse treatment providers statewide, UCLA
asked survey respondents if they had heard of NIATx, implemented procedures, or

2

See https://www.niatx.net/Content/ContentPage.aspx?NID=49

127

communicated with the NIATx national organization.3 Out of 86 responses from the random
sample, 12 programs (14.0%) indicated that they had heard of NIATx while 71 said they had
not, and three declined to answer. Five programs reported having actually implemented
NIATx procedures (5.8%). Two of these programs were part of the Los Angeles County
Process Improvement Pilot Project that will be described later in this chapter, the other three
were not. Three providers (3.5%) indicated that they had been in contact with
representatives from the NIATx organization. Two of these had participated in the pilot
project, the other had not (See Table 7.2).
Table 7.2: Awareness of NIATx (n=86)

% Yes
Have you heard of the Network for the Improvement of
Addiction Treatment (NIATx)?
Has your treatment program ever implemented NIATx
Procedures?
Has your treatment program ever communicated with
representatives from the NIATx organization?

14.0
5.8
3.5

The Los Angeles County Process Improvement Pilot Project4
For the past seven years, UCLA ISAP has evaluated county contracted alcohol and drug
treatment and recovery programs in Los Angeles County through an effort known as the Los
Angeles County Evaluation System (LACES): An Outcomes Reporting Program. Through
LACES, it has become apparent that engagement and retention is a significant challenge in
Los Angeles County. High dropout rates occur early in treatment: approximately 25% of
those scheduled for an assessment appointment fail to appear, and a further 25% of those
assessed and referred for treatment fail to attend. Moreover, in many treatment programs,
25% to 50% of clients drop-out of treatment in the first 30 days. Accordingly, rates of dropout before completion in some outpatient treatment programs are as high as 80%.
The Pacific Southwest Addiction Technology Transfer Center (PSATTC) proposed that the
County of Los Angeles Department of Public Health, Alcohol and Drug Program
Administration (LA County ADPA), the PSATTC, the Center for Substance Abuse
Treatment (CSAT), and the NIATx National Program Office collaborate to conduct a
demonstration project to determine whether the process improvement model promoted by
NIATx could improve treatment retention and completion rates locally.
On March 17, 2005, the PSATTC and LA County ADPA convened a three-hour
Informational Meeting for addiction treatment providers in Los Angeles County to introduce
3

The NIATx National Program Office, led by Dr. David Gustafson, Professor of Industrial Engineering, is
located in the Center for Health Systems Research and Analysis at the University of Wisconsin at Madison.
4
The description of the Los Angeles County Process Improvement Pilot Project procedures and outcomes
included here has been adapted from The Los Angeles County Process Improvement Pilot Project
Implementation Guide and Final Report (Rutkowski, 2007). Stakeholders interested in receiving this
document, which includes additional project-specific supporting materials and details, can request it from Beth
Rutkowski at UCLA ISAP (finnerty@ucla.edu).

128

them to the principles of process improvement. Seventeen Executive Directors (or their
designees) from 14 treatment and recovery programs attended the meeting.
Attendees were provided with an overview of the origin and goals of NIATx, as well as a
description of the principles and key roles of the process improvement model. A Project
Director from PROTOTYPES, a program that had previously received funding for NIATxrelated activities through CSAT’s STAR program, described her program’s experience with
process improvement. The attendees then participated in a group discussion to gauge their
collective level of interest in participating in a pilot of the NIATx program in Los Angeles.
In the weeks immediately following the Informational Meeting, the PSATTC Project
Director contacted all meeting participants to see if there was an interest within their
respective program to participate in a structured pilot project. After several months of crossprogram discussions, planning, and preparation, the Los Angeles County Process
Improvement Pilot Project (hereafter referred to as the “pilot project” or “project”) was
formally launched in November 2005.
Pilot Project Participants
Seven treatment programs (six outpatient and one residential) participated in the pilot
project:
•
•
•
•
•
•
•

Didi Hirsch Community Mental Health Center, Via Avanta (Pacoima)
Los Angeles Centers for Alcohol and Drug Abuse (Santa Fe Springs)
Matrix Institute on Addictions – San Fernando Valley
Matrix Institute on Addictions – West Los Angeles
Social Model Recovery Systems, Inc. (Covina)
Southern California Alcohol and Drug Programs, Inc. (Downey)
Tarzana Treatment Centers, Inc. (Tarzana)

Three of these agencies treated Prop 36 clients (Social Model Recovery Systems, Southern
California Alcohol and Drug Programs, Inc., and Tarzana Treatment Centers, Inc.).
However, because lessons learned from all seven programs are potentially applicable to Prop
36 treatment, the activities of all seven are discussed below.
Pilot Project Leaders and Co-Sponsors
The Director of LA County ADPA endorsed the project and assigned an Executive Sponsor
and Change Leader who attended pilot project meetings, workshops, and conferences. In
addition, the Executive Sponsor and/or Change Leader participated in monthly conference
calls.
The PSATTC coordinated the logistics of meetings and conference calls, provided technical
assistance to participating programs, co-facilitated program site visits, managed the monthly
flow of project information, and collaborated with the partner programs in planning the
project design and implementing the project activities.

129

The NIATx National Program Office provided technical assistance by contributing faculty
and content for all meetings and conference calls, and co-facilitating the program site visits.
In addition, the National Program Office provided technical assistance to the PSATTC with
regard to the design and revision of the client-specific tracking worksheet.
The Center for Substance Abuse Treatment, through a logistics contract with AFYA Inc.,
made provisions for a maximum of $10,000 ($5,000 each to LA County ADPA and the
PSATTC) to cover logistic costs (staff time and food and beverage charges were not
allowable expenses). Staff time was provided in-kind by the PSATTC, LA County ADPA,
and the NIATx National Program Office.
Project Director
The PSATTC Project Director contributed the equivalent of about .2 FTE during the course
of the 11-month project. This position was critical to the ongoing management of the
project, and without this level of effort there would be little chance of success.
Process Improvement Coach
A Process Improvement Coach was assigned to the project. The Process Improvement
Coach collaborated with (and mentored) the Project Director; helped plan and facilitate all
face-to-face meetings/workshops, and conference calls; and co-facilitated the half-day
program site visits. The Process Improvement Coach contributed approximately 15 days
during the 11 month tenure of the project.
Key Aims of the Pilot Project
Several objectives were originally formulated for the project:
•

•

•
•

To determine whether programs receiving minimal support and no financial
assistance could adopt and use the NIATx process improvement methods to
improve client retention and continuation rates in substance abuse treatment.
To use data collected by the participating programs to determine the degree
to which they are able to improve participation, reduce no-shows, and
increase 30- and 60-day continuation rates in substance abuse treatment.
To assess program commitment to adopting and administratively supporting
the process improvement methodology at the conclusion of the pilot project.
To identify key factors that contribute to project success and components that
can be improved in the future.

The Four Phases of the Pilot Project
Phase 1: Pre-Work (3 months)
Executive Sponsors who were considering participating in the pilot project were invited to
the pilot project Orientation Meeting. The goal of the meeting was to stress the importance
of the CEO/Program Director making a commitment of time and personnel to the project.
The Institute for Healthcare Improvement (2003) suggests such commitment is typically
marked by the following:

130

•
•
•
•
•
•

An aim for improvement is established and overseen by leadership at the
highest level in the organization.
Measures and change strategies are consistent with strategic plan or key
priorities.
Leadership is able to channel program attention to the change process and
results.
The Change Leader has the influence and time to devote to process
improvement.
Direct service staff are engaged in the improvement process.
Program leaders see the business case for the benefits of improvement.

At this three-hour Orientation Meeting, participants were asked to prepare for project
implementation by: (1) assigning an Executive Sponsor who would support the project by
making it a program priority, remove potential barriers, and participate directly when
necessary; (2) assigning a Change Leader who would provide daily leadership, keep the
project organized, and assure that the Change Team is continually working to achieve
improved results; (3) developing a baseline (through the compilation of existing data or
collection of new data) over two months on the following: assessment and first appointment
no-shows, and 30- and 60-day client continuation rates; and (4) conducting an agency walkthrough to identify potential improvements to existing program procedures used in the
assessment, admission, and active phases of the treatment process.
Phase 2: Kick-off Workshop
The pilot project was officially initiated at an all day workshop. The goals of the workshop
were to: (1) build interest and confidence in conducting process improvement projects; (2)
familiarize Change Team members with the concepts underlying a structured improvement
process and the use of rapid cycle change strategies; (3) provide an opportunity to prioritize
improvement needs based on existing data and the experience of conducting a program
“walk-through;” (4) create a quick-start roadmap for initiating service improvements; and (5)
clarify the sequence of planned project activities.
At the workshop, participants were exposed to evidenced-based process improvement
practices, heard case examples from peers, participated in interactive exercises, and
developed a process improvement “quick start roadmap” that could be used to guide the
process improvement project rollout at their respective programs, networked with other
participants, and scheduled future site visits and monthly conference calls.
Phase 3: Change Project Implementation (8 months)
Throughout the pilot project, Change Teams from each participating program conducted one
or more process improvement rapid Plan-Do-Study-Act cycles aimed at reducing no-shows
and improving continuation and completion rates. Pilot organizations were expected to start
a project focused on reducing no-shows or increasing admissions until they achieved at least
a 20% improvement. This 20% figure was a suggested minimum, but participants were
encouraged to choose goals that were ambitious yet also realistic and achievable. Thereafter,
programs were asked to focus their efforts on reducing drop-outs and improving continuation
rates at 30 and/or 60 days following admission. Program site visits were held just prior to
131

the start of change project implementation. Activities during the project implementation
phase included data collection, Monthly Change Leader conference calls, one Executive
Sponsor conference call, and a Change Leader Face-to-Face Meeting. Details of these
activities are described below.
Data Collection: All participating programs collected and submitted a monthly Microsoft
Excel client tracking worksheet (designed by and available from the NIATx National
Program Office) and a Microsoft Word change project reporting form to the Project
Director, which compiled necessary client/program information to monitor progress and
to troubleshoot potential implementation issues.
The Excel worksheet tracked key client dates (e.g. first request for service, intake
appointment, admission, discharge). Embedded formulas throughout the spreadsheet
automatically calculated no-show rates, continuation rates, etc. The worksheet also
generated graphs that programs could print out and use during Change Team meetings to
illustrate the impact of the changes that were being tested.
The MS Word change project reporting form was provided to assist Change Teams in
keeping track of their various change projects throughout the course of the project
implementation period. The form detailed the basic information on the project, details of the
Plan-Do-Study-Act rapid cycle changes, and project outcomes and sustainability plans. See
Appendix 7 for a copy of this form.
Program Site Visits: The Project Director and Process Improvement Coach conducted halfday site visits with each participating program, allowing them to be introduced to the staff
and the facilities of the programs involved in the pilot project. During the site visits, the key
features of process improvement were reviewed, and the data collection tools were
explained. During some visits, the Process Improvement Coach and Project Director were
given the opportunity to participate in a Change Team meeting or to consult on the change
projects being planned and implemented.
One of the major topics covered during each site visit was a detailed review of the change
project reporting form and the client-specific data tracking worksheet. It became apparent
early in the site visit week that the client tracking worksheet needed to be altered to better
meet the needs of the individual programs. Changes were made to the form in early March
and the revised form was communicated via email to the Change Leaders, as well as to the
identified data person (if he/she differed from the Change Leader).
Monthly Change Leader Conference Calls: The project design allowed for a series of
Change Leader conference calls (open to all Change Team members). The Project Director
and Process Improvement Coach shared responsibility for facilitating the calls. The purpose
of the monthly calls was to provide a scheduled “check in time” for the participating
programs to share their experiences to date, ask questions, and receive clarification on the
data reporting forms. It also gave the Project Director the opportunity to share information
on upcoming project events. Four Change Leader conference calls were held throughout the
project implementation phase.

132

Executive Sponsor Conference Call: A single Executive Sponsor conference call occurred
early in the project implementation phase. The purpose of this call was to provide a “check
in time” for the participating programs to make sure that the Executive Sponsors were onboard with the changes that their respective program’s Change Team was planning to
implement. The Executive Sponsors were provided with a brief review of the first Change
Leader conference call that occurred the day before. One question raised was how soon to
discontinue a change if it seems as if it is not working. The Process Improvement Coach
recommended that a change be given enough time to allow a fair test. Sometimes a fair test
is just one week, other times it is a month or more. In general, anywhere from 25-50% of all
changes are aborted, due to a lack of impact. One major benefit of conducting rapid cycle
testing is that it allows changes to be altered quickly and easily.
Change Leader Face-to-Face Meeting: About half-way through the project implementation
phase, a half-day Change Leader Meeting was held. There was 100% participation from the
programs, with several programs choosing to bring multiple members of their Teams,
including Executive Sponsors. Each Change Leader was given a copy of the data from
his/her program (to use for comparison purposes). The Change Teams were encouraged to
keep making progress and also to begin to look at ways to improve their 30- and 60-day
continuation rates in addition to engagement/access.
Phase 4: Completion Conference
In September 2006, the Change Teams from each program were invited to a half-day
Completion Conference. The purpose of the event was: (1) to celebrate the successes of each
Change Team (by reporting on a change project that led to improvements in client
engagement/access and/or retention/continuation); and (2) to share ideas regarding the
continuation of process improvement strategies within Los Angeles County and
sustainability of the current change projects.
At the Completion Conference, each program provided a 15-minute presentation on their
change project experiences. The presentations included descriptive information as well as
actual monthly change data. This was facilitated by four ready-made graphs (covering noshows, admissions, and continuation) included in their MS Excel data sheets that could
easily be incorporated into the presentations. All programs also created a poster to illustrate
the information they shared during their oral presentations.
Post-Project Focus Groups
In July 2007, approximately ten months after the September 2006 Completion conference,
UCLA conducted follow-up focus groups as part of the UCLA evaluation of Prop 36. Focus
groups were conducted with the three programs that had participated in the Pilot Project that
served Prop 36 clients. The goal of these focus group interviews was to discuss the long
term perceptions and effects of participation in the pilot project among Prop 36 programs.
Focus group methods are detailed in Appendix C. Results will be discussed following the
description of the projects below.

Treatment Program Change Projects:
The following information was extracted from presentations given at the project completion
conference, site visit summary notes, monthly Change Leader conference call minutes, and

133

monthly change project reporting forms/client tracking worksheets. The highlights are
organized in alphabetical order, by program name. Because treatment programs presented
their own data, there is some variation in the information available for each program.

Didi Hirsch Community Mental Health Center, Via Avanta
Via Avanta, the only residential treatment program included in the pilot project, provides
treatment to women with children under the age of 5. The program staff uses a therapeutic
community model. At any given time, there are approximately 40 residents (as well as 15
children under the age of 5). The average length of stay is roughly six months (180 days).
The program does not typically serve Prop 36 clients.
The key engagement and retention issues were identified through a variety of methods,
including a program walk-through, baseline data collection, and focus groups with clients
who were in treatment for less than 60 days.
Baseline Data
According to baseline data collected from December 2005 to February 2006, 44% of clients
were discharged within 30 days of treatment entry (that is, 56% continued for at least 30
days).
Key Issues Identified and Actions Taken
Issue #1: New clients (those in treatment for less than two months) felt that the intake
process was too impersonal and that the many rules and responsibilities were very
overwhelming. In addition, too many program staff members were involved in the intake
process.
Rapid Change Cycle #1 (initiated 2/27/06): New clients would not be given any community
responsibilities for the first 15 days in treatment. New clients were encouraged to be selfpaced during their first two weeks in treatment. The change allowed for and a client-driven
orientation to treatment.
Issue #2: New clients did not like having to find someone in the program to be with them at
all times during the first 15 days in treatment (known as “finding” or “calling cover”). The
clients found the process to be humiliating, and felt that it added pressure to the treatment
and recovery process.
Rapid Change Cycle #2 (initiated 4/26/06): New clients would no longer have to be
“covered” by a peer during the first 15 days of treatment.
Issue #3: Because of change #2, “Big Sisters” (clients who were in treatment for 90+ days)
were disengaging from their “Little Sisters.”
Rapid Cycle Change #3 (initiated 5/24/06): The treatment staff employed motivational
enhancements with the Big Sisters. Along with a Change Team member, the Big Sisters
developed a checklist of responsibilities and goals that they would attempt to meet on a
weekly basis. The goals were designed to engage the little sisters in the treatment process.
When a goal was reached, a reward would be given (a group activity such as a movie, coffee
outing to Starbucks, pizza party, ice cream social, etc).
134

Results
In the five months spanning from February 27, 2006 to July 31, 2006, the average 30-day
continuation rate increased. The initial goal was to increase the continuation rate by 20%,
but the Via Avanta Change Team was able to exceed this goal, increasing the continuation
rate by 54% (from 56% to 86%) (See Figure 7.1).

Percentage

Figure 7.1
30 Day Continuation Rate
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

86%
56%

Baseline Period (Jul 05-Feb 06)
n=75

Project Period (Feb 06 - July 06)
n=35

Time Period

A side effect of this improvement was an increase in clients’ average length of stay, from
approximately 125 days (in the five months prior to the project) to 175 days (in the five
months of change project implementation). Additionally, there were about 30 fewer
admissions than usual between March and July, but because the women were staying in
treatment longer, the residential program contracts were maximized.
Lessons Learned
The clients loved the changes. Staff members were slightly hesitant initially to believe that
the changes they made were responsible for the reduced drop-out rate. It had to be brought
to their attention that they were retaining clients in treatment for a longer period of time.
During the completion conference, the program emphasized several lessons learned:
•
•
•
•
•
•

The Change Leader and Team must stay focused and committed to improvement;
New ideas can get lost if not implemented;
Improvement is a process that does not end – it is ongoing;
Statistics help with evaluation and accountability;
Positive changes need to be reinforced as soon as possible; and
The project resulted in a program culture change;

135

Los Angeles Centers for Alcohol and Drug Abuse
The Los Angeles Centers for Alcohol and Drug Abuse (LA CADA) is a non-profit
community program dedicated to treatment and prevention of substance abuse, HIV/AIDS,
and related problems for Los Angeles County communities. The four LA CADA
departments include outpatient services, Allen House residential services, HIV/AIDS
prevention services, and Family Foundations. The pilot project was implemented in the
adult outpatient program in Santa Fe Springs. This program does not typically serve Prop 36
clients.
Baseline Data
During the baseline period of November-December 2005, 40% of adult clients did not attend
one or more scheduled group or individual sessions. The Change Team set a goal to reduce
the combined individual/group session no-show rate to 30% (a 25% improvement in
attendance).
Key Issues Identified and Actions Taken
Issue #1: Progress in reducing the session no-show rate would be measured by utilizing a
data collection form and reviewing the client attendance rosters maintained by the staff
counselors. Counselors would log their clients’ scheduled visits on a weekly basis. At the
end of each week, they would indicate which sessions were attended, the number of
groups/individual sessions missed, etc.
Rapid Change Cycle #1 (initiated 3/06): The LA CADA Change Team developed a change
project known throughout the program as “Target Attendance.” Several simultaneous
changes were made during a two-month implementation period, including the following: (1)
full color flyers were hung throughout the site to alert clients to the availability of incentives
for perfect attendance; (2) incentives were offered in the form of $5.00 gift cards to various
local merchants, such as Target, Starbucks, AMC Theatres, etc.; (3) pot luck celebrations
were held and (4) monthly pledge cards were given to clients (that included the weekly
outpatient group schedule). By signing the pledge cards, clients were agreeing to strive for
perfect attendance.
Clients were given a month to make up missed sessions. During the potluck lunches, a
recognition ceremony was held for those clients who had perfect attendance and were to
receive an incentive. Clients received more than one incentive if they had 100% attendance
in multiple months. Those who had perfect attendance were invited to choose a gift card
from a fish bowl.
Results
As shown in Figure 7.2, the session no-show rate dropped from 40% at baseline to 28%.
This translates into an improvement of 30%, which exceeded the initial goal of 25%.
Lessons Learned
During the June Change Leader Meeting and September 2006 Completion Conference, the
program communicated the following impressions and lessons learned:
•

It was best to “Keep it simple.”;

136

•
•
•
•
•

It was best to find positive solutions to motivate clients;
Meeting weekly helped to keep everyone informed of changes and ensured
that everyone was on the same page;
The main challenge was gathering data;
Not only were the clients motivated to show up for their group and individual
sessions, but the outpatient counseling team was, as well; and
Revenue increased as a result of increased session attendance.
Figure 7.2
No Show Rates

Percentage No Shows

60%
50%
40%

40%
28%

30%
20%
10%
0%
Baseline Period (Nov-Dec 2005) n=229

Project Period (March-May 2006) n=185

Time Period

Matrix Institute on Addictions, San Fernando Valley
The Matrix Institute provides adult and adolescent outpatient drug and alcohol treatment
services across several locations in the greater Los Angeles area. The adult intensive
outpatient program based in the San Fernando Valley (Tarzana) was staffed by five clinicians
and two part-time administrative assistants. Approximately 85 Prop 36 clients, 25 privatepay clients, and 19-30 social support (continuing care – voluntary, alumni) are served in any
given month at this location. The standard length of the manualized program is two or four
months, depending on insurance or other funding coverage.
Baseline Data
According to baseline data collected in August 2005, the percent of private-pay clients
continuing treatment for at least 60 days was 75%. Because the 60-day continuation rate at
this location was relatively good to begin with, there was not a large margin for
improvement. The desired goal was to increase the 60-day continuation rate to 80% (a 7%
improvement).
Key Issues Identified and Actions Taken
Issue #1: The Change Team was unsure if the clients enrolled in the intensive outpatient
program felt that they had the opportunity to provide their therapists with initial impressions
of their treatment experience, as well as suggestions for what could be done to improve their
overall experience. The Change Team chose a change that would be easy on both staff and
clients to implement.
137

Rapid Change Cycle #1 (initiated 3/01/06): In an early change meeting, the Team decided to
query the clients to see how they felt about their treatment experience thus far, and what
could be done to improve their ongoing experience in the program. The Team decided to
take a non-incentive approach to improving continuation by developing a four-item
questionnaire. Administration of the questionnaire allowed the therapists to have an
additional individual contact and gave their clients a chance to share their feelings about
what was happening to them and how they felt about their treatment experience.
The therapists waited until the clients were in treatment for about three weeks before
administering the questionnaire. The therapists attempted to administer the questionnaire
individually, either in person or on the phone. The four questions were:
•

Is there a particular topic (not in the manual) that you would like to see me
bring in to the relapse prevention group?
As your therapist, what can I do to enhance our relationship in the next few
months?
Since starting Matrix three weeks ago, what have you found most valuable
and least valuable in your treatment experience?
In the next 30 days, what would you like me to pay attention to in your
recovery?

•
•
•

It is important to note that during the course of the pilot project, there were staffing changes.
As a result, the Change Leader took over a case load. It is therefore difficult to determine
whether the improvement in retention was due to the implementation of the questionnaire or
to the change in staffing.
Results
Figure 7.3
60 Day Continuation Rate

Percentage

100%
80%

85%

75%

60%
40%
20%
0%
Baseline Period (Aug 05)

Project Period (Mar 06-July 06) n=24

Time Period

138

The average 60-day continuation rate from March to July 2006 was 85%, compared to 75%
in the baseline period. This translates into an improvement in continuation of 13%, which
exceeded the original goal of 7% (see Figure 7.3).
Lessons Learned
The clients appreciated the opportunity to provide feedback on topics for the treatment
manual. On average, the clients’ favorite question was “what topic would you like added to
the manual?” The change project was straightforward and easy to implement.
At the September 2006 Completion Conference, the Executive Sponsor outlined the change
team’s impressions of participating in the pilot project. One obstacle was staff turnover,
which made continuity difficult. In addition, it was sometimes hard to establish a regular
meeting time that would work with everyone’s schedules. Lastly, having to wait two months
before observing if the change was impacting the 60-day continuation rate was difficult. But
overall, the staff enjoyed the team approach used in the pilot project. They felt that they had
a direct impact on the changes that were being implemented, including longer retention in
treatment.

Matrix Institute on Addictions, West Los Angeles
The adult intensive outpatient program based in West Los Angeles, California was staffed by
three full-time therapists who treat anywhere from 12 to 20 private-pay clients per month.
The treatment design is the same as the one described above for the San Fernando Valley
program.
Baseline Data
According to baseline data collected in August 2005, the percent of private-pay clients
staying in treatment for at least 60 days was 75%. Because the 60-day continuation rate at
this location was relatively good to begin with, there was not a large margin for
improvement. The desired goal was to increase the 60-day continuation rate to 80% (a 7%
improvement).
Key Issues Identified and Actions Taken
Issue #1: Clients enrolled in treatment at this particular location primarily participated in
group sessions (i.e., three group sessions per week). It was a challenge to get them to show
up to their individual counseling sessions.
Rapid Change Cycle #1 (initiated 03/01/06): Prior to beginning the pilot project, the
members of the Change Team met weekly for group supervision. This was a natural time to
incorporate a weekly change meeting.
To help clients keep track of their individual session appointments, the Change Team
provided clients with a 4-session appointment card (stapled to the back of their therapist’s
card) that included the dates of the four individual session appointments scheduled
throughout the first 60 days of treatment.
The clients would first hear about the appointment cards during their evaluation (assessment)
session, so they would be prepared to schedule their individual appointments with their
therapist when the time came to do so. At the first individual session, the therapist gave each
139

client an appointment card and scheduled three subsequent individual sessions (for a total of
four individual sessions). In addition, the appointments were listed in the program’s
scheduling book.
There was a brief period of adjustment early on in the process, but by the second month of
the change cycle, the Team got used to giving the appointment cards to every client. An
unexpected result of this change project was that the intake coordinator received fewer calls
from clients who were not sure when their next individual session was scheduled.
Results
The average 60-day continuation rate throughout the course of the project was 61%.
Although the Change Team was able to increase the 60-day continuation rate within the
active project period (from 56% in March 2006 to 64% in June 2006, and then again from
57% in May 2006 to 64% in June 2006), the average 60-day continuation rate of 61% was
not an improvement over the baseline continuation rate of 75% (See Figure 7.4).
Figure 7.4
60 Day Continuation Rate
100%

Percentage

80%

75%
61%

60%
40%
20%
0%
Baseline Period (August 05)

Project Period (March 06-June 06) n=62

Time Period

The Change Team was not sure whether results were affected by client insurance plans that
would only pay for a limited number of sessions. The Team reported that the sometimes
clients were not able to make their appointments right away, and it was easy for them to “slip
through the cracks.”
Lessons Learned
Throughout the project, the Change Team reported that when looking at the data, they saw
things that they otherwise might not have, and they thought it would be a good idea to
continue to track attendance behavior by clients in the future (applying the tracking to other
treatment groups). The team also felt the procedures they learned were very helpful.

Social Model Recovery Systems, Inc.
Social Model Recovery Systems, Inc. (SMRS) is a 12-Step based program that incorporates
role modeling by residents and program staff and peer support to achieve goals. A total of

140

six treatment and recovery programs (2 residential and 4 outpatient) are operated throughout
Los Angeles and Orange counties. The pilot project was implemented in the Prop 36
outpatient program in Covina.
Baseline Data
According to baseline data collected between November 16, 2005 and January 16, 2006,
37% of clients who were referred by the Prop 36 Community Assessment and Service
Centers (CASCs) for an intake appointment failed to show for the appointment. The Change
Team sought to reduce the intake no-show rate to 30% (for an approximate 20%
improvement). Additionally, the 30-day continuation rate at baseline was 65%. Once the
assessment no-show change project was under way, the Change Team sought to increase the
30-day continuation rate to 85% (for an approximate 30% improvement).
Key Issues Identified and Actions Taken
Issue #1: The Change Team discovered that potential Prop 36 clients were being referred by
the Community Assessment Service Centers (CASCs), but were not making it to SMRS for
their intake appointment. They decided to focus on reducing these intake no-shows through
the use of various strategies.
Rapid Change Cycle #1 (initiated February/March 2006): The first change project involved
building rapport with the CASCs that were referring Prop 36 clients to SMRS. The Change
Team decided to interact directly with these CASCs by visiting them to introduce themselves
to the assessors. They provided the assessors with bus tokens, pamphlets offering
information regarding available services at SMRS, and SMRS business cards to pass out to
potential clients. They also invited the CASC assessors to visit the program. When clients
came in for their intake appointment, the Change Team members asked them if they received
a business card from the CASC to verify that the CASCs were doing what Change Team
members had asked them to do.
Rapid Change Cycle #2 (initiated March/April 2006): In addition to building relationships
with the CASC assessors and potential clients, the Change Team decided to move to same
day (or next day) intakes to reduce waiting time. They hoped that clients would be more
likely to appear for their appointment and enter treatment if there was a shorter amount of
time between scheduling and appearing for their appointment. The program manager
generally did all of the intakes, but achieved some flexibility by providing training to other
staff members so they could act as a back up if the program manager got overbooked or was
otherwise unavailable and someone needed to be assessed.
Issue #2: Once the Change Team felt that they had the intake no-show issue under control,
they moved on to another problem area. Clients were dropping out of treatment early (within
the first 30 days of treatment), and some participants were showing little or no motivation to
participate fully in treatment.
Rapid Change Cycle #3 (initiated May 2006): The Change Team decided to increase
participation in treatment by offering a variety of incentives. The goal was for clients to
have 100% group/individual session and 12-Step meeting attendance, as well as provide a
negative urinalysis. Each participant who achieved 100% in all areas would receive a

141

department store gift card at the end of the month. The Change Team informed the clients of
the changes and created a tracking chart to monitor attendance.
Once the incentive project was under way, the Change Team determined that they needed to
identify creative ways to raise money to provide the incentives. The Change Team organized
a “Salad Express” that sold salad bar-style lunches to members of the SMRS community for
a “donation” of $4.50. This “Salad Express” raised over $100. An ice cream social raised
approximately $60.00. The money raised from these events went towards purchasing gift
cards for the incentive project.
In July, the Change Team noticed that client attendance decreased briefly in the prior two
weeks. They decided to build up client morale by asking alumni to come together with
current clients to plan a day of fun including a BBQ picnic. In September, the Change Team
planned an outing to attend a taping of a television game show.
Results
No-Show Rates: The average intake no-show rate during the active project period (MarchJuly 2006) was 1.1%, which represented a 97% improvement over the baseline rate of 37%
(see Figure 7.5). This greatly exceeded the desired improvement of 20%.

Percentage No Shows

Figure 7.5
No Show Rates
60%
50%
40%

37%

30%
20%
10%

1%

0%
Baseline Period (November 2005-January
2006)

Project Period (March-July 2006) n=45

Time Period

Notable side benefits resulted from the business card/same day appointment intervention –
the number of CASC referrals increased, and 100% of those who were referred and entered
into treatment attended their first group session (up from 69%). The Change Team felt that
the same day assessment appointments probably had the biggest impact on the number of
referrals. Shortly after the change project began, the program maxed out on the capacity of
their Prop 36 program.
Thirty Day Continuation Rates: The average 30-day continuation rate for May through July
2006 was 75%. This translated into an improvement in continuation of 15% (see Figure 7.6).
With incentives and encouragement, the clients appeared to be more motivated to participate
142

in treatment. In June 2006, four Prop 36 clients had 100% participation and were awarded a
$25 gift card (to Target, Starbucks, movie theatres, or gas stations). Another result of the
incentive project was that the clients were being taught social skills and tools they could use
to interact in a public environment without using alcohol and drugs.
Figure 7.6
30 Day Continuation Rate

Percentage

100%
80%

75%

65%

60%
40%
20%
0%
Baseline Period (November 2005January 2006)

Project Period (May-July 2006)

Time Period

Lessons Learned
The Change Team members reflected on several lessons learned throughout the pilot project:
•

•
•

At first it was difficult to get the Team together due to conflicting
meetings/appointments, varying work schedules, etc. The Team now meets
once a week on Friday afternoons. After initial responses such as “Do we get
any more money,” and “Great, more meetings to attend,” the Change Team
members embraced the project, and worked hard to achieve successes;
Data collection has become a routine part of the team’s schedule; and
The Team needs to stay on top of change strategies or they will be forgotten;

Southern California Alcohol and Drug Programs, Inc.
Southern California Alcohol and Drug Programs, Inc. (SCADP) is a non-profit organization
dedicated to the prevention and treatment of substance abuse and related problems. SCADP
targets underserved and disadvantaged populations, including homeless, victims of domestic
violence, persons living with HIV/AIDS, pregnant and parenting women and their children,
as well at Deaf and Hard of Hearing persons and the criminally-involved. The program
provides 500 residential treatment beds and 1,000+ outpatient counseling slots that serve
over 5,000 men, women, and children each year throughout Los Angeles and Orange
counties. The pilot project was implemented in the Prop 36 outpatient program in Downey.
Baseline Data
The baseline data collected for November-December 2005 indicated a no-show rate for
intake appointments of 57%. In addition, the early discharge rate (within the first 30 days)
was 28%. The Change Team set a goal to decrease the intake no-show rate to 35% (for a
39% improvement).
143

A separate set of baseline data collected during March-April 2006 for the incentive project
revealed that 22% of clients met all 4 criteria for Prop 36 compliance (100% attendance in
group sessions, 12-step meeting participation, providing drug tests as scheduled, and 100%
attendance in all individual counseling appointments.) The Change Team was curious if
incentives would lead to any increase the percentage of clients who were in full compliance
with the program requirements (a goal was not chosen).
Key Issues Identified and Actions Taken
Issue #1: Initially, the Change Team thought they would focus on a perceived attendance
problem. But it turned out that there was an approximate 80% attendance rate, even
through the holiday season. However, through the collection of baseline data, they found
out they had a problem with missed intake appointments.
Rapid Change Cycle #1 (initiated in 3/06): As was the case with SMRS, Prop 36 clients are
referred to SCADP by CASCs. Prior to the pilot project, the CASC assessors would call
SCADP and whoever answered the phone would schedule the intake appointment.
Generally, there was little or no contact between the counselor and client prior to the intake
appointment. The Change Team decided that the Prop 36 counselors would talk with the
potential client when the CASC called to schedule the intake appointment. When the CASC
assessor called, the counselor would introduce him/herself, tell the prospective client a little
about the outpatient program, and ask the client if he/she had any specific needs that should
be addressed during treatment. Motivational interviewing-type strategies were utilized by
the counselors on the calls.
Issue #2: Once the Change Team decided to sustain the initial change, they moved on to
another identified problem area – client retention.
Rapid Change Cycle (initiated in 5/06):
The Change Team decided to provide incentives for their clients to encourage greater
participation in the various components of treatment. Incentives (in the form of a $5 gift
card) would be provided to clients who met the following four criteria: (1) attended all
individual and (2) group treatment groups for a month, (3) came in for all scheduled drug
tests, and (4) attended the required number of 12-Step meetings. The incentives would be
presented during group sessions so that the group could acknowledge and congratulate each
individual client for his/her effort.
Results
Reducing No Shows: The average assessment no-show rate between March and July 2006
was 14.4%, which represented a 75% improvement over the baseline no-show rate of
57% (see Figure 7.7).
The Change Team learned early on that the initial contact that the counselor made with the
prospective client helped to decrease the no-show rate. A side benefit of the initial counselor
contact was an increased likelihood that a client would complete the assessment appointment
and eventually be admitted for treatment. In February, 18 of 21 clients who completed an
assessment appointment enrolled in treatment (86%); in March, the percentage increased to
100%.

144

Figure 7.7
No Show Rates

Percentage No Shows

60%

57%

50%
40%
30%
14%

20%
10%
0%
Baseline Period (November-December 2005)

Project Period (March-July 2006) n=52

Time Period
In May, the counselors stopped talking with prospective clients when the CASC called and
the Change Team stopped meeting on a weekly basis as other job responsibilities got in the
way. The result was an increase of the no-show rate from 11% to 33%. Because of this setback, the Change Team recommitted to these activities.
Increasing Compliance with Prop 36: The two-month average percentage of clients meeting
all four criteria was 28%, which translated into a 27% improvement over baseline (see
Figure 7.8). Even though a higher percentage of clients met all four criteria in May and June
than during the baseline period (Mar-Apr), the Change Team decided that having to meet all
4 criteria to receive $5.00 was too strict. Instead, in future months, clients would receive a
$5.00 gift card for each criterion they met (with the possibility of earning up to $20.00 in gift
cards) each month.
In July 2006, 37% of clients met one or more criteria and received at least one $5.00 gift
card; and in August 2006, 35% of clients met one or more criteria and received at least one
$5.00 gift card. The criterion that yielded the best results was drug testing (45% of clients in
July and 58% of clients in August took their drug tests as required).
In September, the treatment group with the highest attendance rates was treated to a raffle
party. All Prop 36 clients were eligible for at least one raffle ticket. Additionally, each
client who met one or more of the four incentive criteria was given a raffle ticket (one for
each criterion they met). The winner of the grand prize, a bicycle, was a 30 year-old male
client who took public transportation to treatment.
Lessons Learned
Change Team members reflected on a variety of lessons learned, including the importance of
data collection and of being open-minded to change. In addition, the Team realized how
important it was to have consistent change meetings and open lines of communication. The
Change Team started off well, but got sidetracked for a while when the Change Leader was
moved to another project. The Team realized what was happening and worked together to
correct the problem.
145

Figure 7.8
Met all Four Criteria
60%

Percentage

50%
40%
30%

28%
22%

20%
10%
0%
Baseline Period (March - April 2006)

Project Period (May-June 2006)

Time Period

The Change Team also learned that by working together, department processes could be
improved. This allowed for staff members to be more efficient. Data collection was timeconsuming and tedious, but very enlightening.
SCADP has had an incentive fund available for some time now. By participating in the pilot
project, the outpatient program staff members were able to make better use of these funds,
with little financial impact. Lastly, listening to other programs’ Change Project ideas led to
new ideas for process improvement within SCADP.

Tarzana Treatment Centers, Inc.
Tarzana Treatment Centers, Inc. provides behavioral healthcare and treatment services in
Los Angeles County through a continuum of integrated alcohol and drug addiction treatment,
education, mental health, medical detoxification, and residential rehabilitation for
teens/youth, and adults. They also provide outpatient services, sober living housing,
continuing care, HIV/AIDS services, Prop 36, family medical care, women’s services, family
counseling, domestic violence intervention, anger management, and community education
services. The pilot project was implemented in the adult intensive outpatient (level 3) Prop
36 program in Tarzana.
Baseline Data
Baseline data collected during November-December 2005 indicated that 61% of clients
remained in treatment for more than 30 days. The Change Team decided to try and increase
their 30-day continuation rate by 50% (from 61% to 92%).
Issues Identified and Actions Taken
Issue #1: The main issue identified was a high drop-out rate in the first 30 days of treatment.
Rapid Change Cycle #1 (initiated 3/17/06): The first project the Change Team initiated was
the development and implementation of two questions that staff would ask clients during the
intake appointment. The Change Team wondered if demonstrating an interest in potential

146

treatment road blocks could help to increase retention further down the line. The two
questions were as follows:
1. What is going on in your life that would prevent you from coming to treatment?
2. What resources or support do you need to come to treatment?
Client responses to the first question indicated that work schedule, transportation, traffic, and
anxiety were stumbling blocks. Responses to the second question ranged from “nothing” to
needing bus tokens or other transportation to treatment. The Team decided to create a new
intervention plan, continue to gather data, and explore options to address the identified
concerns.
Rapid Change Cycle #2 (initiated 4/14/06): Next, the Change Team decided to coordinate a
group or individual session with the client’s primary counselor within 24-48 hours of
admission. When early drop-outs decreased, the Team decided to monitor and confirm oneon-one session contact with counselor and perhaps consider tokens and vouchers for session
attendance.
Rapid Change Cycle #3 (5/05/06): As a continuation of change cycle #2, the Change Team
decided to confirm that the clients’ sessions with their primary counselor were taking place
within the first 24 to 48 hours of treatment enrollment. This involved tracking the dates
between admission and first post-admission treatment session. The team found that the dropout rate continued to decrease, and the retention rate increased.
Rapid Change Cycle #4 (6/02/06): Lastly, the Team decided that counselors should be
proactive in monitoring patient needs. The Change Team received approval from the
Executive Sponsor to initiate a future change project involving incentives.
Results
The average 30-day continuation rate throughout the project was 94%. This represented a
54% improvement over the baseline rate of 61% (which exceeded their goal of a 50%
improvement) (See Figure 7.9).
In May and June 2006, the average time between admission and the first post-admission
treatment session was 1.9 days, which was within the 24-48 hour period that the Change
Team had designated as acceptable during rapid change cycle #3.
Like other programs that participated in the pilot project, TTC experienced a slow down in
Prop 36 admissions due to over-utilization of the Prop 36 funds earlier in the contract year.
Lessons Learned
As an incidental result of the pilot project, the TTC Change Team realized that some of their
admission criteria were not clear. During the project, a few clients were discharged because
they needed a higher level of care. These clients would have benefited from a more thorough
assessment, so that they were properly placed from the beginning.

147

Figure 7.9
30 Day Continuation Rate
94%

100%

Percentage

80%
61%

60%
40%
20%
0%
Baseline Period (November-December Project Period (March-July 2006) n=35
2005)

Time Period

Overall Results
Data from the six outpatient/intensive outpatient programs was combined to illustrate
aggregate improvements in intake no-show and 30-day continuation rates. The average noshow rate (from March to July) was 6.8%, which represented an 80% improvement over the
baseline rate of 34% (See Figure 7.10).

Percentage No Shows

Figure 7.10
Aggregate No Show Rates
60%
50%
40%

34%

30%
20%

11%

10%

7.10%

5.70%

2.40%

7.70%

6.80%

0%
Baseline

March
2006

April 2006 May 2006 June 2006 July 2006

Average

Time Period

The average 30-day continuation rate (from March to July) was 75%, which represented a
6% improvement over the baseline continuation rate of 71%. See Figure 7.11.
It is possible that the lack of significant improvement in 30-continuation rates was due to
insufficient time to test multiple strategies. Several programs chose first to focus on

148

improving their no-show rates to assessments/intakes or their individual/group session
attendance. After demonstrating significant improvement in reducing no-shows to intake or
increasing attendance to individual/group sessions, there may not have been enough time to
adequately test innovations that could improve 30-day continuation prior to the conclusion of
the pilot project. Other NIATx projects have successfully increased retention (McCarty et
al., 2007).
Figure 7.11
Aggregate 30 Day Continuation Rates

Percentage

100%
80%

92%

85%
71%

73%

75%

73%
54%

60%
40%
20%
0%

Baseline

March
2006

April 2006 May 2006 June 2006 July 2006

Average

Time Period

Overall Lessons Learned
The following lists of lessons learned were generated from two complementary sources:
•

•

Throughout the project, participants were given the opportunity to share their
impressions of process improvement and of pilot project implementation.
Observations and feedback collected over the course of the project were summarized
by participants, the Project Director, the Process Improvement Coach, and the ADPA
Executive Sponsor at the end of the project.
Second, approximately ten months after the September 2006 Completion conference,
UCLA conducted three follow-up focus groups with the treatment programs that
served Prop 36 clients. The goal of these interviews was to discuss the longer term
effects of participation in the pilot project, assess whether programs were able to
sustain progress made during the project, and re-evaluate perceptions of the project
given the passage of time. In particular, these interviews were focused on perceived
keys to implementation, overcoming barriers, and sustainability.

Conclusions from both of these sources are included on the following pages.

149

Keys to Implementation
Pilot Project: During the September 2006 Completion Conference, the Process
Improvement Coach presented a list of what he believed to be seven cumulative lessons
learned. The lessons were:
•
•
•
•
•
•
•

Seeing things from the client’s perspective can be helpful.
Multiple improvements can be made in a short period of time.
Process improvement can motivate staff and clients when good results occur.
The results surpassed the initial objectives/expectations.
Simple improvements yield big dividends.
Using data can actually be helpful.
There is great value in “sticking with it” (i.e. sustaining effort and keeping
communication flowing).

The following list includes feedback from pilot project participants recorded throughout the
project, as well as impressions from the Project Director and Process Improvement Coach:
•
•

•
•

•

•

•

Staff members who deal with patients can generate innovative ideas.
The structure of the project allowed for a collaborative, not competitive
process. The program representatives were respectful of one another the
entire time.
The mentoring provided by the Process Improvement Coach and Project
Director was helpful.
Management’s attitude and enthusiasm regarding the project goes a long way
in ensuring demonstrable results. It became apparent early on that the
programs that did not have executive buy-in faltered along the way and had
to work extra hard to make modest improvements in engagement and
retention.
Key variables that predicted which programs would be successful and which
would struggle were: commitment of the Executive Sponsor, degree of
interest from the Change Leader, and the quality of questions that were
asked.
In general, the more invested the Executive Sponsor was in the changes being
made, the more likely it was that the changes would be sustained in the long
run.
It would have been productive to spend more time providing instructions on
how to complete and maintain the monthly client tracking forms and change
project reporting forms. Because the client-specific tracking form needed to
be altered after the site visits (due to technical difficulties), the changes had
to be communicated to the sites via e-mail and phone alerts, as opposed to in
person. In future projects, it is recommended that participating programs be
provided with an in-person training on how to maintain the MS Excel client
tracking form and the MS Word change project reporting form.

150

•
•

One of the most valuable aspects of participation in the pilot project was the
opportunity to interact regularly with the County, as well as other programs.
The pilot project was an innovative way for participants to look at
procedures, ask questions, and see what would work to increase retention. It
was simple yet effective.

Follow-up Focus Groups:
•

•

•

•

•
•
•

Two groups noted that the data allowed them to see where they started and
ended, and that this was motivating. Participants from the other program said
they had already been tracking data. However participants in all groups
expressed satisfaction in finding that small changes could have a substantial
impact on outcomes.
Participants said it produced valuable change in organizational culture,
allowing changes and providing an avenue for asking why things were being
done in certain ways, which empowered staff. “We were frustrated anyways,
and here was a program that was able to help us vent our frustration and, and
offer us suggestions on how we can improve this process.”
All groups noted that buy-in was the key to implementation. Groups noted
that buy-in from the Executive Sponsor was critical. Change Teams felt they
had a mandate. One group noted that by working together as a team they
were able to push through changes more effectively. Another noted that the
data was important in getting buy in from staff and the director.
Two of the groups stressed the importance of having the change team
meeting at least weekly, and that all of the members must be on the same
team.
Having a team that gets along well, has complementary interests and skills,
and is able to communicate well was seen as crucial.
All groups said the conference calls, meetings, and guidance from the Project
Director and Process Improvement Coach were helpful.
One group noted that having ready-made databases and charts was very
helpful.

Overcoming Barriers
Pilot Project:
•

The project design did not provide enough individual coaching for the
Change Leaders. Although Change Leaders were invited to contact the
project team anytime they had questions or were having difficulty, such
contact occurred relatively infrequently. In future projects incorporation of
one monthly telephone contact with each Change Leader to allow for a
discussion of issues that are unique to the particular program is
recommended, in addition to the monthly conference calls with all programs.
If this kind of monthly call is implemented, it would be useful to distribute an
agenda prior to the call, take minutes, and distribute them to the participants.
Doing so will help track progress and make for more accurate reporting.
151

•

The conference calls were not as well attended as originally hoped. The
highest number of participating programs on the calls was five out of seven.
Typically three or four programs were represented. Those programs that
were inconsistent in their reporting of data were also the least likely to
participate in the conference calls.

Follow-up Focus Groups:
•

•

One barrier cited was time and personnel resources needed to staff the project
and keep it running. One group dealt with this by making sure everyone
came to meetings prepared, which made meetings short and efficient.
Another program noted that initially the data aspect was time consuming, but
once developed was easy to maintain. The other program noted that the
process was easier than expected and the team saw immediate results.
Some frustration was also expressed at not being able to make changes that
would have required extra funding (e.g. incentive programs, transportation),
which limited their options. In most cases, changes that required funding
were not pursued. However, in one program, previously unknown program
funds that could be used for incentives were discovered, and another program
began using incentives that were either donated from nearby merchants or
were essentially free (certificates handed out in front of peers and family).

Suggestions for Sustainability
Pilot Project: The LA ADPA Executive Sponsor facilitated a discussion by asking,
“Where do we go from here?” Recommendations drawn from attendee comments are
listed below:
•

•
•
•
•
•
•

Funds should be made available to programs to offer incentives to clients and
contractual language should be changed to earmark a certain amount of
money for contingency management/motivational incentives-type activities.
Positive process improvement results should be rewarded.
A process improvement section should be added to the ADP website as well
as a link the to the NIATx website.
Counties should help programs secure donations/technology improvements,
such as computer software and hardware.
Participating program representatives could serve as coaches for new
programs.
There needs to be effective communication between counties and their
contracted provider programs.
Counties should focus recurrent lecture series on process improvement.

152

Follow-up Focus Groups:
•
•

•

One program suggested technical assistance to get a team started, perhaps
from teams that have participated in a pilot project before.
One program reported that having someone come visit the program at the
start, then come back later to follow up was useful. They also noted that
having someone from outside the program regularly check in to ask how
things are going really helps to keep the program on track and prevent the
program from letting things drop.
Two groups noted that having someone always available to answer questions
(like the Project Director and Process Improvement Coach were during the
pilot project) on a continuing basis would be helpful.

Ongoing Efforts
Follow-up Focus Groups:
•

•

•

•

All three programs reported maintaining some or all of the changes that had been
made during the pilot project. One program, however, reported that one change
“has been up and down” due to a counselor shortage.
All programs reported that the pilot project had changed their general perspective
on treatment. For example in one group, a participant noted that “you don’t go
back” and there was a consensus that as a result of the pilot project they now
think of the client as the customer, which represented a shift in thinking. One
participant in the group also remarked that the pilot project made him/her aware
that one doesn’t have to be in charge of the program to make changes.
Previously, the participant would come up with ideas and think “if I were in
charge…” but the ideas would die there.
Some changes continue to be made, but they were not always being
systematically tested. One program reported that the change team and Plan-DoStudy-Act cycle were still in place. Another reported that the cycle was “put to
bed” when the project ended. The other group said there is no formal change
team anymore but that similar activities are continuing “in a different way from
the change team.”
Staff movement and turnover tended to disperse the Change Teams, representing
a major barrier to sustained effort.
o One program reported that all of the change team members are no
longer there. The Program Manager said that she has not been able to
train the new staff member on his new job and aspects of the process
improvement methods, although he does do some tracking of client
information.
o In another program, a participant reported trying to keep the cycle
alive in terms of collecting data, but s/he did not report making or
testing any new changes after the pilot project. The participant
reported that s/he is collecting data for his/her “own use.”

153

•

o Partly as a result of the success of the pilot project, one change team
member was actually promoted to become Director of a residential
facility. Ironically this meant the success of the change team’s efforts
resulted in its own dispersal.
All three groups reported that the changes had spread to other parts of their
program.
o The participant who had been promoted to Director of a residential
facility reported taking some NIATx processes to the new program,
but said s/he did not have the staff to collect and enter the data
systematically there. When asked how changes are evaluated, the
participant reported observing everything in person and to see how
things are working. Participants from the original facility also assist.
For example, the participant implemented a change at the residential
facility that required staff to say their names when they answered the
phone. When staff from the unit that participated in the pilot project
called the residential facility they would check on whether staff were
following these instructions and reported their experiences to the new
director. The Director also noted that he was trying to keep the
changes small and implement them one at a time so he could evaluate
them, consistent with NIATx procedures.
o Another program reported that the successful pilot project changes
that had started in one portion of the facility had spread to other parts.
o Another program reported that a “new change team” had been created
at the program’s residential facility. Group members reported “I
think that’s helped them” but they were “not really sure” if they were
using the same NIATx steps used by the outpatient program that
participated in the pilot project.

Next Steps
Pilot Project: The LA County Executive Sponsor asked completion conference participants
what they would change if another pilot project were to be implemented. Responses
included:
•
•
•
•

Slow things down a bit – it was too much, too fast this time.
Include a “trial” period to test out strategies to improve continuation rates.
Develop a community to keep changes going.
Continue the walk-through process – it is a very good initiation to process
improvement.

In November 2007, the PSATTC and LA County ADPA commenced phase II of the Los
Angeles County Process Improvement Pilot Project. A total of 12 agencies have been
recruited to participate, including 10 new agencies and two agencies that participated in
phase I. These two agencies wish to spread the changes made during the phase I project to
another program within their agency. The phase II program objectives and key activities are
nearly identical to those from the phase I project, although a few changes were implemented

154

based on feedback received from phase I participants. The phase II pilot project will end in
October 2008.
Also in November 2007 the PSATTC sponsored three daylong trainings on improving client
access and retention in treatment as part of the ongoing California Addiction Training and
Education Series (CATES). The daylong trainings will be followed by six months of followup technical assistance conference calls, facilitated by the trainer and PSATTC Associate
Director.

Discussion
Future process improvement efforts in the substance abuse treatment field may be informed
by related efforts to apply continuous quality improvement in health and mental health care,
which have resulted in mixed results (for a review see Shortell et al, 1998). Four dimensions
are crucial for significant organizational improvement to occur: strategic, cultural, technical,
and structural. If any of these are missing, the result is likely to be little or no impact
(O’Brien, et al., 1995; Shortell et al, 1998).
Strategic
This dimension refers to processes that are strategically most important to the organization.
Failure on this dimension means the organization is wasting its energy on peripheral, less
strategically important activities, and as a result the effort will have little to no impact on
important activities. This problem typically arises from an inability to select goals that
would clearly fit into the organization’s strategic priorities, and failure to make quality
improvement a central part of organizational planning.
Cultural
This dimension refers to the underlying beliefs, values, norms, and behaviors of the
organization that inhibit or support improvement work. Failure on this dimension means the
improvement is not appraised, celebrated, or rewarded. The result is small, temporary effects
and quick backsliding. This problem typically arises when organizations look inward to the
needs of their workers rather than outward to the needs of their customers, when personnel
resist working as teams, or when improvement is perceived as primarily a cost-control
mechanism.
Technical
This dimension refers to the training and information support system issues. Failure along
this dimension means people are not sufficiently trained and/or supporting data analysis are
inadequate. This results in frustration and false starts. This problem typically arises from
lack of team-based, problem-focused training, insufficient provision for ongoing training and
upgrading of skills, and inadequate or nonexistent information systems.
Structural
This dimension refers to the presence or absence of mechanisms to facilitate learning and to
disseminate “best practices” throughout the organization via task forces, committees,
steering councils, communication, etc. Failure along this dimension results in an inability to
capture lessons learned and spread it throughout the organization. This typically arises from

155

failure to take full advantage of the resources of organization-wide steering councils or
similar groups, and lack of alignment between budgeting and planning systems.
By extension, process improvement applications are most likely to be successful when all of
the dimensions above facilitate success. The following conditions provide the best
probability of success (Shortell, et al., 1998):
•
•

•

When carefully focused on areas of real importance to the organization and
addressed with clearly formulated interventions.
When the organization is ready for change and has prepared itself by
appointing capable leadership, creating relationships of trust with physicians,
and developing adequate information systems.
When there is a conducive external environment relative to beneficial
regulatory, payment policy, and competitive factors.

Applied to Prop 36, this suggests that while trainings and limited-term projects have value,
ultimately permanent incentives and infrastructure must be provided to continually support
and encourage continuing improvement, or the efforts will falter.

Conclusions and Recommendations
Overall, the pilot project resulted in successful changes at all participating programs, and at
the end of the project each treatment program generally maintained their progress.
Furthermore, ten months after the end of the pilot project programs had generally kept in
place the changes that they had made during the project. However, results were mixed on
continuing process improvement. In some cases it essentially slowed to a halt, while in some
cases the process had at least partially spread to other portions of the program.
The alteration and spread of the NIATx methods that occurred after the pilot project ended is
simultaneously encouraging and unsettling. Based on both pilot project feedback and
reviews of the application of improvement efforts in health care, it is clear that process
improvement efforts can be ineffective, frustrating, and/or a waste of resources if
implemented incorrectly. Therefore, while there is no evidence of any negative effects in the
follow-up interviews, there is a danger that the spread of partial and altered NIATx methods
could eventually lead to negative results, perhaps as a sense of being a ‘waste of time.’
In order to maintain the fidelity of the process improvement system, a controlled roll-out is
recommended. Participants in the pilot program reported that guidance from the Project
Director and Process Improvement Coach were instrumental in their success, and that in
particular technical assistance with data collection was a key element. Without similar or
greater levels of support, it is unlikely that other programs would have the success seen in
the pilot project. Since cost was a limiting factor primarily at the initiation stage of the
process improvement, it would be helpful for funding agencies to provide funding for such
activities, with the understanding that the program would sustain change processes after the
initial period of funding.
In order to create lasting and complete improvement efforts it will be critical to create a
permanent infrastructure to support program staff, much in the way the Project Director and

156

Process Improvement Coach supported participants during the pilot project. The
establishment of a permanent process improvement center that can answer questions,
organize conference calls, and monitor programs that are participating in future process
improvement efforts is highly recommended. Such permanent infrastructure can work to
facilitate new efforts such as the pilot project described in this chapter, prevent deterioration
of process improvement efforts in existing participants, and increase the chances of success
where inevitable “spread” of the process occurs to new programs.
Since change is facilitated by endorsement and support of organizational leadership, ADP
and county lead agencies should take leading roles in ensuring that such efforts be expanded
and removing barriers. The potential value of extending these methods to other portions of
the system (e.g. court, probation, parole) should also be explored.
It may also be useful for ADP and county agencies to create incentives for programs that
successfully implement and maintain NIATx methods, whether through performance based
contracting or other means.
When properly and fully implemented, with proper levels of support, the pilot project and
relevant literature have demonstrated that NIATx process improvement methods can produce
substantial benefits at relatively low cost. The success of these methods in reducing noshow rates and increasing treatment continuation show promise in addressing two key areas
of concern in Prop 36.

References
Berwick, D.M. (1989). Continuous improvement as an ideal in health care. New
England Journal of Medicine, 320, 53-56.
Institute for Healthcare Improvement. (2003). The Breakthrough Series: IHI’s
Collaborative Model for Achieving Breakthrough Improvement (IHI Innovation
Series White Paper). Boston: Institute for Healthcare Improvement.
Laffel, G, & Blumenthal, D. (1989). The case for using industrial quality management
science in health care organizations. JAMA, 262, 2869-73.
Langley, G., Nolan, K., Nolan, T., Norman, C., & Provost, L. (1996). The model for
improvement. In The improvement guide: A practical approach to enhancing
organizational performance (3–11). San Francisco: Jossey-Bass Publishers.
McCarty, D., Gustafson, D.H., Wisdom, J.P., Ford, J, Choi, D., Molfenter, T., Capoccia,
V., & Cotter F. (2007). The Network for the Improvement of Addiction Treatment
(NIATx): Enhancing Access and Retention. Drug and Alcohol Dependence, 88,
138-145.
O'Brien, J.L., Shortell, S.M., Hughes, E.F., Foster, R.W., Carman, J.M., Boerstler, H., &
O’Connor, E.J. (1995). An integrative model for organization-wide quality
improvement: lessons from the field. Quality Management in Health Care, 3, 19-30.
Shortell, S.M., Bennett, C.L., & Byck, G.R. (1998). Assessing the impact of continuous
quality improvement on clinical practice: what it will take to accelerate progress.
The Milbank Quarterly, 76, 593-624, 510.

157

158

Chapter 8: Narcotic Treatment Programs
Bradley T. Conner, Ph.D.
The National Consensus Development Panel on Effective Medical Treatment of Opiate
Addiction unequivocally states that narcotic replacement maintenance as part of a
comprehensive narcotic treatment program (NTP) is the most effective means of treating
opioid dependence. Methadone is the most widely used replacement opioid in NTP,
however, buprenorphine has been approved for use as a maintenance medication. The
beneficial outcomes of NTP's far exceed those associated with the treatment of opioid
dependence using detoxification, residential, or outpatient treatment modalities.
California Alcohol and Drug Data System data indicate that, between July 2005 and June
2006, 10,992 individuals, or 6% of all drug treatment clients, were receiving methadone
maintenance. However, methadone maintenance, methadone detoxification, and
buprenorphine maintenance were used infrequently among Prop 36 participants whose
primary drug problem was with an opioid. Likely due to such ‘mismatched’ placements,
treatment completion was lower and treatment duration was shorter for opioid users than
for users of other drugs.
Despite immense research evidence supporting the utility of NTP for reducing drug use
and crime among opioid addicts, public policies, especially anti-NTP attitudes within the
criminal justice community, have hampered the use of NTP in Prop 36. Educators may
need to be more sensitive to ideological differences of opinion not due entirely to a lack
of knowledge, as opposition to NTP exists even after dissemination of significant
research evidence supporting its effectiveness. In this case further education may not
affect change. Targeted education that first collects information regarding the specific
opposition to NTP may be more effective in these instances.
While NTP may not be appropriate for every opioid-dependent Prop 36 participant, it is
an important treatment tool. Buprenorphine may be an attractive alternative NTP
medication for counties that do not currently have NTP available, are unwilling or unable
to open a methadone clinic, or are looking for inventive and cost-effective ways of
implementing NTP in their county. Dosages of both medications should be closely
monitored, accompanying ancillary services should be mandatory, and buprenorphine
should be available as an alternative, however, UCLA continues to urge each county to
make some form of NTP available to Prop 36 participants whose primary drug of choice
is heroin or another opioid.
Epidemiologic studies indicate that opioid dependence in the United States affects
approximately 800,000 people each year (i.e., Office of National Drug Control Policy, 2003).
Opioid abuse and dependence affects people from all segments of American society, as well
as their families and communities. According to the 2004 National Survey on Drug Use and
Health, 31.8 million Americans reported use of prescription opioids for non-medical
purposes in their lifetime and 3.1 million reported use of heroin in their lifetime (Substance
Abuse and Mental Health Services Administration [SAMSHA], 2005). Researchers have
159

estimated the costs of nationwide abuse and dependence of opioids at $21 billion annually
(Mark et al., 2001). In 2006, treatment centers had more than 466,000 admissions for heroin
use (SAMSHA, 2007). Additionally, opiates other than heroin (non-heroin opiates) were the
primary substance of abuse for 51,000 substance abuse treatment admissions (SAMSHA,
2006). Survey data indicate that as many as 12% of the residents of California reported
lifetime non-medical use of prescribed pain relievers (Wright et al., 2007). Additionally,
CADDS data from July 2005 to June 2006 indicate that 10,992, or 6%, of all clients in
treatment for substance abuse, were in methadone maintenance treatment. Within the Prop
36 population, 385, or 0.85% of all clients admitted to drug treatment during the same time
frame received methadone maintenance, even though 3, 167, or 7.53% reported that heroin
or some other opioid was their primary drug of choice.
Two full µ-opioid agonist medications, methadone and Levo-Alpha-Acetylmethadol
(LAAM), and one partial µ-opioid agonist medication, buprenorphine, have the approval of
the U.S. Food and Drug Administration (USFDA) to be used as narcotic replacement
medications for detoxification and maintenance treatment of opioid use disorders. The Drug
Enforcement Administration (DEA) has listed Buprenorphine as a Schedule III drug whereas
methadone and LAAM are Schedule II drugs (United States Department of Justice, Drug
Enforcement Administration, 2007). Schedule III drugs have an accepted medical use and
less potential for abuse or dependence than Schedules I and II drugs. Schedule III drugs are
available only by prescription, though control of wholesale distribution is somewhat less
stringent than Schedule I and II drugs. Prescriptions for Schedule III drugs may be refilled
up to five times within a six month period. Conversely, Schedule II drugs have a high
tendency for abuse and can produce dependency with chronic use. These drugs may have an
accepted medical use and are only available by prescription. Distribution is carefully
controlled and monitored by the DEA. Schedule II drugs are also subject to production
quotas set by the DEA. As a result, these drugs require more stringent records and storage
procedures than drugs listed on Schedules III and IV, however the DEA has imposed similar
records and storage procedures for buprenorphine, though it is a Schedule III substance.
All three medications have been shown to be effective in the treatment of opioid dependence.
Research on the efficacy and effectiveness of NTP has been on going since the 1950s.
Joseph and colleagues (2000) provide a comprehensive review of the research on methadone
maintenance. Longshore and colleagues (2005) and Anglin and colleagues (2007a & b)
provide results of a randomized clinical trial and a comprehensive review of the research on
LAAM maintenance. Ling and colleagues (1998) provide the results of a randomized
clinical trial of buprenorphine maintenance.
The National Consensus Development Panel on Effective Medical Treatment of Opiate
Addiction unequivocally stated that narcotic replacement maintenance (i.e., methadone
maintenance) as part of a comprehensive narcotic treatment program is the most effective
means of treating opioid dependence (NIH Consensus Development Program, 1997). Of
these medications, methadone is the most widely used. Methadone was first developed in
Germany prior to World War II as an analgesia and first used as a treatment for opiate
dependence in the 1950s (Joseph et al., 2000). LAAM, a longer-acting medication than
methadone, was approved for treating opioid dependence in 1993 (USFDA, 1993), however,
manufacture of the medication was discontinued in 2003 (USFDA, 2003). Both methadone
160

and LAAM, according to Federal (SAMSHA, 2001) and State (California Health and Safety
Code 11839-11839.22) regulations, must be administered under very specific conditions and
in highly controlled environments. More recently, buprenorphine has also been approved to
treat opioid dependence (USFDA, 2002). However, because buprenorphine is a Schedule III
drug, it has approval to be delivered through a doctor’s office, rather than a licensed clinic, as
long as the doctor has a valid license to prescribe Schedule III controlled substances (this
qualification is discussed in detail later in this chapter). Federal regulations also allow
certified methadone maintenance programs to prescribe buprenorphine (SAMSHA, 2003),
though participants that receive buprenorphine through a methadone clinic must meet the
typical federal and state requirements for patients who attend these clinics, which eliminates
some of the benefits of using a Schedule III substance rather than a Schedule II substance.

Narcotic Treatment Programs
Narcotic Treatment Programs (NTP) typically treat opioid dependence using two different
paradigms: Detoxification, which is the administration of a substitute opioid for a specified
amount of time (typically 10, 14 or 30 days) starting with a large dosage and tapering the
dosage amount until it reaches zero. The goal is abstinence from opioid use. The basic
rationale of maintenance treatment, however, comes from medical, public health, and harm
reduction perspectives. The underlying principles are that some people are simply unable to
stop using opioids, due in part to physiological changes in the brain that are relatively
permanent, and that both the individual and society will benefit if these individuals are
switched from using illicit drugs to using legal drugs obtained from physicians and
sanctioned treatment clinics. Under the maintenance treatment paradigm there is no defined
treatment cessation date, treatment is ongoing and only ends at the patient’s request, if the
patient excessively violates regulations or clinic policies, or if the patient is unable to pay
and has no access to public funds. The field has reached consensus that maintenance
treatment is the most effective treatment for heroin dependence (American Methadone
Treatment Association, Inc., 2004; Mathias, 1997; National Institute on Drug Abuse, 1999;
NIH Consensus Development Program, 1997), as such this chapter will focus on
maintenance rather than on detoxification.
The primary goals when administering maintenance medications are to:
•
•
•
•
•

relieve narcotic craving
suppress opioid withdrawal syndrome for 24–36 hours
block the effects of administered heroin
develop tolerance to the euphoria, sedation, or other narcotic effects of opioid
medications which impair day-to-day functioning, emotional responses, or
perception while improving functional status
develop tolerance to the analgesic properties of the medications

Individuals receiving methadone in an outpatient clinic typically visit the clinic on a daily
basis to receive their medication. Under its original design the individuals would stay at the
clinic and participate in ancillary services such as drug testing, individual therapy, group
counseling, and vocational training. While this is not always the case in the current funding
era, ancillary services should be a fundamental part of any NTP for optimal benefits to be
achieved. NTP programs that also provide buprenorphine must meet these same conditions.

161

However, individuals receiving buprenorphine from a certified physician see a doctor at the
doctor’s office to obtain a prescription. Appointments can range in frequency from once per
week to once per month, with the typical cycle being one doctor’s appointment every 13
days. Clients are not required to receive ancillary services.
In the state of California, NTPs may be paid for privately or publicly. To assist with
payment for services rendered to individuals who are unable to pay federal, state, and local
funds are distributed to the NTPs through county and direct provider contracts. Funding
sources also include Medi-Cal, and third- party payers such as private insurance companies.

NTP in Prop 36
UCLA collected information on NTP practices in Prop 36 from a survey of court
administrators (see Appendix A). Analysis of the data from the 27 court administrators who
responded suggested that the role of the court in assigning individuals to receive NTP varies
widely across the state. On this survey administrators were asked “Did the court ever assign
Prop 36 opiate users to [NTP]? (…methadone maintenance, for example, not detoxification
only)”. Responses indicated that the court assigned Prop 36 opiate users to NTP in 48.1%
(13) of the counties across the state. Table 8.1 presents the breakdown of court criteria for
placement in NTP. The 10 court administrators who endorsed “Other” gave various reasons
for these placements, including if the initial assessor recommended placement in NTP, if the
client was already in a NTP, or if the client requested NTP.
Table 8.1: Court Criteria for NTP Placement
Number
Only if drug free treatment was unsuccessful*
3
Only if drug free treatment was unavailable
1
As the first option for treating opiate users†
1
Other
10

Percentage
24.1%
7.7%
7.7%
76.9%

* One county endorsed both “only if unsuccessful” and “only if unavailable”
† One county endorsed both “only if unsuccessful” and “other”

Table 8.2: Court Criteria for NTP Non-Placement
Number
*
Not offered to Prop 36 offenders by county policy
4
Narcotic Treatment is unavailable in the county†
4
‡
Philosophical opposition to Narcotic Treatment
1
Other
8

Percentage
33.3%
33.3%
8.3%
66.7%

* One county endorsed “not offered by county policy” “unavailable in county” and “other”
† One county endorsed “not offered by county policy” and “other”
‡ One county endorsed “philosophical opposition” and “other”

Table 8.2 presents the breakdown of court reasons for not placing individuals in NTP. For
the 44.4% (12) that reported not placing clients in NTP, 8 endorsed “Other” and reported
various reasons for not placing clients in NTP including no opioid users in their counties, no
requests for NTP services, just started offering services, or no money to pay for the services.
As has been reported in the previous years’ reports from UCLA, methadone and
buprenorphine were used infrequently in Prop 36. Across the first 5 years of Prop 36, clients
162

who reported an opioid as their primary drug of choice received NTP at 9.9%, 12.7%,
12.9%, 16.0%, and 16.5%. The increase in NTP placements across the years occurred
primarily in the area of methadone detoxification as noted in Chapter 2. In contrast, across
the same years, individuals seeking treatment for opioid use disorders outside of the criminal
justice system have received NTP between 75% and 85% of the time. Figure 8.1 presents
the trends of NTP use by referral source across the first 5 years of Prop 36.

Percentage

100

Figure 8.1
Percentage of Opioid Users Receiving NTP by
Year and Referral Source
85.1

83.1

81.4

80.1

80

75.5

60
40
20

9.7 13.5

12.7 13.4

12.9 13.4

16.1 14.7

16.5

10.8

0
01/02

02/03

03/04

04/05

05/06

Prop 36 Year
Prop 36 Referral

Non-Prop 36 CJ Referral

Individual Referral

Note: Data are from CADDS and include all treatment admissions for Prop 36 probation or parole referrals,
non-Prop 36 Criminal Justice (CJ) referrals, and all non-criminal justice referrals (including self referrals) from
July 1, 2001 to June 30, 2006.

In 2007, UCLA also collected information from a random sampling of Prop 36 treatment
providers (see Appendix B). A total of 91 treatment providers completed the surveys, which
included the item “What percentage of your Prop 36 clients were opiate users at treatment
entry (e.g., heroin, oxycodone, morphine)?” The majority of the programs (78 or 85.7%)
reported that some portion of their clients’ primary drug of choice was an opioid1. The
average (mean) percentage of clients per program was 23.3 (standard deviation 28.1).
Approximately 46% of the programs (44) reported that some or all of these clients were
receiving NTP.
Since Prop 36’s inception, most opioid-using clients participating in Prop 36 were placed in
outpatient drug-free programs. Follow-up analyses conducted by UCLA in a previous
evaluation showed significant differences in treatment outcomes for Prop 36 clients who
received NTP compared with those who did not (Hawken et al., 2007). Opioid users had the
lowest completion rates (26.4%) when compared to users of all other drugs in Prop 36.
Treatment duration was also shorter for opioid users than for users of other drugs. More
1

UCLA oversampled methadone maintenance clinics to ensure representation in the sample. Of the 84
programs that responded to the survey, 6 reported that they were methadone maintenance clinics only.

163

specifically, 71% of opiate-using SACPA clients placed in NTP had a satisfactory treatment
completion compared with 52% of clients who were not placed into NTP. Opiate-using Prop
36 clients who were placed in NTP were significantly more likely to be in compliance with
the treatment provisions of their Prop 36 probation than those placed in other treatment
modalities. NTP clients also had significantly fewer arrests (13% fewer) during a 30 month
follow-up period. NTP clients had significantly fewer drug arrests (an average of 1.1
compared to 1.3 arrests per offender). NTP clients also have significantly fewer property
arrests (an average of 0.2 compared with 0.3 arrests per offender). This data has led UCLA
to the conclusion that opioid users’ performance in Prop 36 will improve significantly if NTP
is made more available.
NTP in the Offender Treatment Program
The Substance Abuse Offender Treatment Program (OTP) was established in Fiscal Year
2006-2007 per Health and Safety Code Division 10.10, Chapter 75, Statutes of 2006
(Assembly Bill 1808). The primary goal of OTP is to enhance the outcomes and
accountability of Prop 36. The OTP statute authorized ADP to distribute appropriated state
general funds to counties that demonstrate a commitment of county matching funds at a ratio
9:1 OTP to county match. ADP outlined a list of goals and strategies that they wanted the
counties to focus on which was informed by recommendations from UCLA. One of these
was to increase NTP availability for treatment of opiate dependent offenders who wish to
receive it.
Thirty-nine counties submitted applications for OTP funding. UCLA coded the applications
which detailed how the requested funds would be used. Approximately 31% of the counties
(12) specified increasing NTP access as one of their goals and detailed their strategies. The
counties varied in size, in the intensity of services they wanted to add, and in the allocation
of the funds towards increases NTP services. The primary strategy for increasing access to
NTP was to increase the number of treatment slots allocated for NTP. The average
allocation amount was $68,178, ranging from $5000 to $185,671. This indicates that ADP
and many counties see the need to increase NTP services available in Prop 36. What is
unclear from the OTP process is whether those that did not indicate increasing NTP
availability thought that they had sufficient NTP resources, did not offer NTP, or used their
OTP funds on other strategies.
Barriers to NTP Utilization
As part of the data collected for this evaluation, UCLA conducted focus groups wherein
participants were asked to discuss their implementation of Prop 36.
One of the more interesting aspects about NTP, which was pointed out a number of times in
the focus groups, is that NTP is the one treatment modality that Prop 36 clients are legally
able to refuse; it is illegal to force an individual to take a medication. This means that Judges
are not able to order someone to a methadone clinic in the same manner that they are able to
order someone to outpatient drug-free treatment, if the individual requests placement in Prop
36. Additionally, NTP is the only treatment in which either a Schedule II or Schedule III
drug is used to treat a drug use disorder. These aspects immediately set NTP apart from the
other treatment modalities that are available and fuel the formation of barriers to its use.

164

Focus group participants reported other common barriers to NTP utilization in Prop 36, for
example:
•

In some counties, the board of supervisors had not approved the use of NTP

•

Some stakeholders do not “believe” the empirical evidence supporting NTP
because of anecdotal evidence that they have collected that contradicts its
effectiveness, such as:
o Most of the people they have met on NTP state that they would
prefer not to be on NTP
o They have been told that it is harder to stop using methadone than
it is to stop using heroin
o They have seen individuals on maintenance treatment who look
“high”

•

Some assessors do not believe in replacing one drug (i.e., heroin) with
another drug (i.e., methadone or buprenorphine) so don’t refer clients to NTP

•

Judges are opposed to NTP so do not allow offenders who come into their
courts already in an NTP program to continue

•

NTP does not fit into Prop 36 which operates in an abstinence model

•

There are no NTP services available in their county or the services that are
available are inappropriate for Prop 36 because they do not offer ancillary
services such as group or individual therapy

•

The county does not have a “heroin problem”

•

The county does not have funding to open a methadone clinic

•

The clinics that exist don’t offer ancillary services, they are not “real”
treatment, just dispensaries

Additionally, interestingly, some stakeholders noted that they were tired of hearing
researchers and NTP providers tout the strengths of NTP, for example, one stakeholder
stated that he was “sick of having methadone shoved down [his] throat”.
There are other factors impacting NTP use in Prop 36, such as perceptions of NTP providers
that the increased workload associated with Prop 36 requirements and Prop 36-specific
contracting issues are too much to make obtaining a Prop 36 treatment contract worthwhile.
Recommendations
Overcoming Barriers to NTP
The primary mechanisms used to date to address barriers to NTP use in Prop 36 have largely
been educational. Through focus groups, UCLA learned that the Judges College usually
offers education on the uses of NRT. One judge indicated that this education changed his
perspective. Other instances of education include seminars offered at the annual Making It
Work conference. This conference is designed to bring stakeholders together to discuss

165

ways to improve Prop 36. Other educational instances include handouts distributed by the
California Opioid Maintenance Providers (COMP) and ADP.
Researchers, treatment providers, and drug treatment advocacy groups also seek to educate
judges and other stakeholders about the uses of NTP (for examples see American
Association for the Treatment of Opioid Dependence, 2006, 2007 and Hora, 2004). These
resources meet with limited success, as they often approach education about NTP by simply
presenting empirical evidence regarding the effectiveness of the treatment.
While UCLA acknowledges that the primary reason for the low rates of assignment to NTP
may be due to lack of education regarding the benefits of such treatment, UCLA is
concerned that the issue is larger than simply being unfamiliar with this type of treatment.
There is ample evidence that supports that many of the individuals responsible for drug
treatment placement under Prop 36, such as judges and assessors, have philosophical
positions in which they oppose the use of narcotics to treat dependence and/or are against the
use of long-term maintenance treatment (in addition to focus group data, for examples of
barriers to NTP in general see Rich et al., 2005). Educators may need to be more sensitive to
ideological differences of opinion that are not due to a lack of knowledge, but because, even
in possession of the research evidence, the person is opposed to the use of NTP for other
reasons. In this case further education may not change the person’s point of view.
Targeted Education
While education is a valuable tool in addressing the barriers associated with increasing NTP
use in Prop 36, the medication must be well designed and targeted to the needs of the
problem. Much of the design of research materials should start much as UCLA did, by
asking stakeholders to define their opposition to NTP. This way researchers and educators
can target specific barriers in specific counties. This also allows the stakeholders to have a
voice. This may lead to stakeholders not feeling like things are being forced upon them.
This may include bringing successfully maintained individuals to educational seminars.

One educational approach may be to target perspectives of people opposed to the use of
NTP. For example, many of the participants in the focus groups indicated that they know
many people on methadone that state that they want to get off of methadone. This is a valid
point, it is likely that most people do not want to have to attend a clinic on a daily basis to
ingest a medication that often has side-effects. However, the likely outcome for many opioid
users who are properly maintained if they stop using methadone is a return to dependence on
the opioid. There is ample research evidence that suggests this would be the case (a detailed
discussion of the chronic illness perspective of drug dependence see McClellan and
colleagues 2000 for a detailed review of drug treatment careers see Hser et al., 1997). So the
better question to ask maintained individuals dependent on opioids is “Do you want to be
maintained on methadone or using heroin?” In this case their answer may be different.
Comparison Studies
In addition to research evidence that exists in the field and the specific results reported
above, evaluation studies comparing outcomes within Prop 36 could, if supportive that NTP
improves outcomes, lead to changing positions and increasing the use of NTP in Prop 36. It
should be easy to compare outcomes between NTP and outpatient drug-free treatment within

166

Prop 36, as there are NTP facilities currently treating Prop 36 participants. This comparison
could also be conducted at the county level, analyzing outcome differences between counties
that use NTP at high rates versus those that use NTP at low rates or don’t use NTP at all.
When comparing at both the program and county level, it will be important to account or
control for differences in ancillary services, as these likely differ across these levels of
analysis. Differences in placement rates and in ancillary services could prove to be very
informative.
Proper Dosage
One of the common statements in the focus groups was that people maintained on methadone
often look as if they are “high on drugs”. Methadone is designed to allow people to function
in daily life. If an individual appears high, they should be referred to the doctor overseeing
medication administration to determine if they are receiving too high of a dose.
Use of Buprenorphine in Prop 36
Currently there are relatively few instances in which buprenorphine is being used
successfully in Prop 36. Though buprenorphine is not currently certified as a reimbursable
medication under California’s Drug-Medi-Cal policy, the California Department of Alcohol
and Drug Programs has identified other methods by which buprenorphine can be paid for as
part of Prop 36: “Suboxone is an allowable [Substance Abuse and Crime Prevention Act] or
OTP expenditure when it is prescribed as part of the [Substance Abuse and Crime Prevention
Act] client's treatment plan through a licensed and certified treatment program” and “In the
event a private physician is prescribing Suboxone as identified in a [Substance Abuse and
Crime Prevention Act] client's treatment plan, the physician would need to be affiliated with
a certified or licensed treatment provider in order for the Suboxone to be an allowable
expense.”

According to federal regulations, licensed physicians (either an M.D. or D.O.) must meet one
or more of the following criteria to qualify for a waiver to prescribe buprenorphine and other
Schedule III drugs under the Drug Addiction Treatment Act (DATA) of 2000 (Public Law
106-310):
•
•
•
•

The physician holds a subspecialty board certification in addiction psychiatry
from the American Board of Medical Specialties.
The physician holds an addiction certification from the American Society of
Addiction Medicine.
The physician holds a subspecialty board certification in addiction medicine
from the American Osteopathic Association.
The physician has, with respect to the treatment and management of opioidaddicted patients, completed not less than eight hours of training (through
classroom situations, seminars at professional society meetings, electronic
communications, or otherwise) that is provided by the American Society of
Addiction Medicine, the American Academy of Addiction Psychiatry, the
American Medical Association, the American Osteopathic Association, the
American Psychiatric Association, or any other organization that the
Secretary determines is appropriate for purposes of this subclause.

167

•

•

•

The physician has participated as an investigator in one or more clinical trials
leading to the approval of a narcotic drug in schedule III, IV, or V for
maintenance or detoxification treatment, as demonstrated by a statement
submitted to the Secretary by the sponsor of such approved drug.
The physician has such other training or experience as the State medical
licensing board (of the State in which the physician will provide maintenance
or detoxification treatment) considers to demonstrate the ability of the
physician to treat and manage opioid-addicted patients.
The physician has such other training or experience as the Secretary
considers to demonstrate the ability of the physician to treat and manage
opioid-addicted patients. Any criteria of the Secretary under this subclause
shall be established by regulation. Any such criteria are effective only for 3
years after the date on which the criteria are promulgated, but may be
extended for such additional discrete 3-year periods as the Secretary
considers appropriate for purposes of this subclause. Such an extension of
criteria may only be effectuated through a statement published in the Federal
Register by the Secretary during the 30-day period preceding the end of the
3-year period involved.

With regard specifically to buprenorphine, DATA 2000 was amended in December 2006,
specifying that an individual physician may have up to 30 patients on buprenorphine at any
one time for the first year but that after one year of certification, the physician may submit a
request to increase this quota to treat up to 100 patients on buprenorphine.
Narcotic Treatment Programs Access in Every County
Though there is opposition to its use by some Prop 36 stakeholders, UCLA is recommending
that an option to receive NTP be available in all 58 counties across the state as part of Prop
36. Current systems for dealing with clients who enter Prop 36 who are already taking a
narcotic replacement medication are unacceptable (i.e., forcing the individual to stop taking
the medication, having the individual drive to a different county to obtain the medication).
Changes in laws that allow a Schedule III drug to be used for maintenance treatment make it
unnecessary to open a methadone clinic in each county to satisfy this recommendation. This
may mean using some inventive solutions, however. For example, each county could
employ or contract with a physician who is eligible to prescribe Suboxone (4:1
buprenorphine - naloxone). Suboxone contains naloxone, which is an opioid antagonist that
can cause opioid withdrawal symptoms if it is injected. This greatly reduces abuse and
diversion liability associated with buprenorphine alone (Subutex) and methadone (for review
see Raisch et al., 2002).

The most effective NTP programs offer a variety of services, not just medication
administration. ADP may want to develop minimum standards for NTP programs that
receive Prop 36 contracts. This would insure the provision of appropriate ancillary services
in combination with the medication.
Conclusions
Despite the unquestionable utility of maintenance medications for reducing drug use and
crime among those dependent on opioids, public policies, anti-NTP attitudes within the
168

criminal justice system and limited access to NTP have hampered the use of NTP, especially
for offenders. While NTP may not be the appropriate treatment for every Prop 36 participant
who reports an opioid as their primary drug, it is an important tool in the treatment of opioid
dependence. UCLA continues to urge each county to make some form of NTP available to
Prop 36 participants whose primary drug of choice is an opioid, such as heroin or
Oxycodone. Buprenorphine may be an attractive alternative NTP medication for counties
that do not currently have NTP available, are unwilling or unable to open a methadone clinic,
or are looking for inventive and cost-effective ways of implementing NTP in their county.
References
American Association for the Treatment of Opioid Dependence (2006). Why Methadone
Works. Accessed at: http://www.aatod.org/fact_methadone.html.
American Association for the Treatment of Opioid Dependence (2007). Drug Court Fact
Sheet: Methadone Maintenance and Other Pharmacotherapeutic Interventions in the
Treatment of Opioid Dependence. Accessed at:
http://www.aatod.org/fact_drug_court.html.
American Methadone Treatment Association, Inc. Fact sheet: Why methadone treatment
works. American Association for the Treatment of Opioid Dependence. Accessed
at: http://www.aatod.org/fact_methadone.html.
Anglin M.D., Conner B.T., Annon J., & Longshore D. (2007a). Levo-AlphaAcetylmethadol (LAAM) versus Methadone Maintenance: 1-Year Treatment
Retention, Outcomes, and Status. Addiction, 102, 1432-1442.
Anglin M.D., Conner B.T., & Longshore D. (2007b). Levo-Alpha-Acetylmethadol
(LAAM) versus Methadone Maintenance: Six-month Post Treatment Outcomes and
Behavior Patterns. Manuscript under review for publication.
California Health and Safety Code 11839-11839.22. Accessed at:
http://www.leginfo.ca.gov/cgibin/waisgate?WAISdocID=76538928933+3+0+0&WAISaction=retrieve.
Hawken A., Anglin M.D., & Conner, B.T. (2007). Treatment differences. Evaluation of
the Substance Abuse and Crime Prevention Act Final Report (pp. 81-93).
University of California, Los Angeles.
Hora P.F. (2004). Trading one drug for another? What drug treatment court
professionals need to learn about opioid replacement therapy. Journal of
Maintenance in the Addictions, 2, 71-76.
Hser Y.I., Anglin M.D., Grella C., Longshore D., & Prendergast M.L. (1997). Drug
treatment careers: A conceptual framework and existing research findings. Journal
of Substance Abuse Treatment, 14, 543-558.
Joseph H., Stancliff S., & Langrod J. (2000). Methadone maintenance treatment
(MMT): A review of the historical and clinical issues. The Mount Sinai Journal of
Medicine, 67, 347-364.

169

Ling W., Charuvastra C., Collins J.F., Batki S., Brown L.S. Jr., Kintaudi P., Wesson
D.R., McNicholas L., Tusel D.J., Malkerneker U., Renner J.A. Jr., Santos E.,
Casadonte P., Fye C., Stine S., Wang R.I., & Segal D. (1998). Buprenorphine
maintenance treatment of opiate dependence: a multicenter, randomized clinical trial.
Addiction, 93, 475-486.
Longshore D., Annon J., Anglin M.D., & Rawson R.A. (2005). Levo-AlphaAcetylmethadol (LAAM) versus methadone: Treatment retention and opiate use.
Addiction, 100, 1131-1139.
Mark T., Woody G.E., Juday T., & Kleber H.D. (2001). The societal costs of heroin
addiction. Drug and Alcohol Dependence, 61, 195-206.
Mathias, R. (1997). NIH panel calls for expanded methadone treatment for heroin
addiction. NIDA Notes, 12.
McClellan A.T., Lewis D.C., O’Brien C.P., & Kleber H.D. (2000). Drug dependence, a
chronic mental illness: Implications for treatment, insurance, and outcomes
evaluations. Journal of the American Medical Association, 284, 1689-1695.
National Institute on Drug Abuse (1999). Principles of drug addiction treatment (NIH
Publication No. 99-4180). Washington DC: National Institutes of Health.
National Institutes of Health Consensus Development Program (1997). Effective
Medical Treatment of Opiate Addiction. Accessed at:
http://consensus.nih.gov/1997/1998TreatOpiateAddiction108html.htm.
Office of National Drug Control Policy Drug Policy Information Clearinghouse. Heroin
Fact Sheet June 2003. Accessed at:
http://www.whitehousedrugpolicy.gov/drugfact/heroin/index.html.
Public Law 106-310: Children's Health Act of 2000. Accessed at:
http://frwebgate.access.gpo.gov/cgibin/getdoc.cgi?dbname=106_cong_public_laws&docid=f:publ310.106.
Raisch D.W., Fye C.L., Boardman K.D., & Sather M.R. (2002). Opioid dependence
treatment, including buprenorphine/naloxone. The Annals of Pharmacotherapy, 36,
312-321.
Rich J.D., McKenzie M., Shield D.C., Wolf F.A., Key R.G., Poshkus M., & Clarke J.
(2005). Linkage with methadone treatment upon release from incarceration: A
promising opportunity. Journal of Addictive Diseases, 24, 49-59.
Substance Abuse and Mental Health Services Administration (2001). Opioid drugs in
maintenance and detoxification treatment of opiate addiction; Final rule. Federal
Register, 66, 4075-4102. Accessed at: http://frwebgate.access.gpo.gov/cgibin/getdoc.cgi?dbname=2001_register&docid=01-723-filed.
Substance Abuse and Mental Health Services Administration (2003). Opioid drugs in
maintenance and detoxification treatment of opiate addiction; addition of
buprenorphine and buprenorphine combination to list of approved opioid treatment
medications. Federal Register, 68, 27937-27938. Accessed at:
http://buprenorphine.samhsa.gov/InterimFinalRule-05-22-2003.pdf.

170

Substance Abuse and Mental Health Services Administration (2005). Treatment Episode
Data Set: 1993-2003. National Admissions to Substance Abuse Treatment Services.
Rockville, MD.
Substance Abuse and Mental Health Services Administration (2006). Non-Heroin Opiate
Admissions: 2003. Drug and Alcohol Services Information System, 16. Rockville,
MD.
Substance Abuse and Mental Health Services Administration (2007). Results from the
2006 National Survey on Drug Use and Health: National Findings (Office of Applied
Studies, NSDUH Series H-32, DHHS Publication No. SMA 07-4293). Rockville,
MD.
United States Department of Justice Drug Enforcement Administration Office of
Diversion Control Drug & Chemical Evaluation Section (2007). Lists of:
Scheduling Actions Controlled Substances Regulated Chemicals. Accessed at:
http://www.deadiversion.usdoj.gov/schedules/orangebook2007.pdf.
United States Food and Drug Administration Talk Paper T93-36 (August 3, 1993). FDA
Announces LAAM Approved to Treat Drug Dependence. Accessed at:
http://www.fda.gov/bbs/topics/ANSWERS/ANS00517.html.
United States Food and Drug Administration Talk Paper T02-38 (October 8, 2002). FDA
Announces Subutex and Suboxone Approved to Treat Opiate Dependence.
Accessed at: http://www.fda.gov/bbs/topics/ANSWERS/2002/ANS01165.html.
United States Food and Drug Administration Product Discontinuation Notice NDC 00543649-63 (August 23, 2003). Letter from Roxane. Accessed at:
http://www.fda.gov/cder/drug/shortages/orlaam.htm.
Wright, D., Sathe, N., & Spagnola, K. (2007). State Estimates of Substance Use from
the 2004–2005 National Surveys on Drug Use and Health (DHHS Publication No.
SMA 07-4235, NSDUH Series H-31). Rockville, MD: Substance Abuse and Mental
Health Services Administration, Office of Applied Studies.

171

172

Chapter 9: Residential Treatment
Angela Hawken, Ph.D.

There was a large statewide increase in the number of clients presenting for drug
treatment as a result of Prop 36 and a large increase in the number of heavy-using clients
in need of more-intensive treatment services. But due to funding constraints and other
barriers to treatment expansion (such as zoning laws and community resistance), the
increase in demand was met largely by expanding less-expensive treatment options,
namely, outpatient care. In many cases, poor performance in outpatient treatment became
an informal criterion for admittance to residential treatment.
The over-reliance on outpatient treatment affected Prop 36 treatment and criminal justice
outcomes. Our analysis of heavy-using Prop 36 treatment clients showed that those who
entered residential treatment were twice as likely to complete their treatment program
(and therefore be in compliance with the terms of their Prop 36 probation), than those
heavy-user clients who were placed into outpatient care. Criminal justice outcomes were
also at issue. Arrest and conviction rates were higher for those heavy-user clients placed
into outpatient treatment. The recidivism differential between residential and outpatient
care was largest for those heavy-user clients presenting to treatment with
methamphetamine as their primary drug problem. This suggests that, from a criminal
justice and public safety perspective, heavy-user methamphetamine clients should be
prioritized for residential care.
Concerns regarding the limited use of residential treatment were raised across stakeholder
groups in the UCLA 2007 focus groups and surveys. Common themes from the
perspectives collected were: concerns regarding the limited availability of residential
treatment slots; the “fail-outpatient first” approach; insufficient lengths of stay in
residential care; lack of sober-living facilities and continuity of care services, and issues
regarding the lack of funding available to reimburse for Prop 36 residential beds and
after-care services. Sixty percent of counties who submitted requests for funding via a
new funding mechanism, Offender Treatment Program (OTP), planned to use these funds
to expand residential care. Many stakeholders noted the importance of OTP funds to pay
for Prop 36 residential beds, and there were concerns regarding the implications of Prop
36 funding cuts for the future of residential placement.
Counties and treatment providers are vulnerable to funding decisions made at the state
level. Only 32% of Prop 36 treatment providers reported that they were able to secure
supplemental funding to facilitate operations as inflation eroded the buying power of Prop
36’s flat budget over 5 years. More recently, this included access to OTP funds which
have since been reduced.
Cuts to Prop 36 funding will pose significant challenges to residential treatment
provision; in the face of inflation erosion and budget pressures, counties are likely to cut
back, rather than expand, existing treatment services. Our analysis indicates that the
resulting implications for residential treatment will impact Prop 36 treatment completion
rates and criminal justice outcomes.
173

The implementation of Prop 36 resulted in a substantial increase in demand for treatment
services across the state, especially for higher severity users. The number of clients referred
to treatment through the criminal justice system doubled statewide in Prop 36’s first two
years, with a large increase in the number of heavy-users. The increase in treatment demand
due to Prop 36 was met largely by expanding services that could be provided at low
additional cost, such as outpatient drug free services. Expanding more-intensive and moreexpensive programs, such as residential care, posed a greater challenge.
Treatment capacity to accommodate Prop 36 clients has increased across the state but lagged
behind the demand for residential placements for heavy users (daily users of an illicit drug at
the time of treatment entry). Many counties maintain residential placement waiting lists and
many clients who might otherwise have been placed into residential care were placed into
outpatient programs instead. The UCLA 2006 Final Evaluation report showed monthly
trends in treatment placement for heavy users entering treatment through the criminal justice
system; the probability of a residential placement fell significantly (for court referrals in
general and specifically, for high-severity court referrals) following Prop 36 implementation.
Appropriate treatment placement of Prop 36 clients is a concern, as treatment is more
effective when clients are matched with services according to the severity of their addiction
and related problems. Research has shown, for example, that the treatment setting and type
of program a drug treatment client is initially placed into makes a significant difference in
the duration of treatment and outcomes (McClellan, 2003). Even though more intensive
services are typically associated with higher costs, research has shown that matching
referrals to the appropriate level of care can lead to lower treatment costs over the course of a
client’s treatment history due to the longer treatment duration or repeated treatment episodes
needed by mismatched clients to achieve desirable outcomes (Sharon et al., 2003).
There is no clear evidence that treatment modality (residential versus outpatient) matters for
the typical client, but research has shown that those with higher drug use and less social
support, or psychiatric comorbidity do better in residential treatment than in outpatient
treatment (Gastfriend & McClellan, 1997; Magura et al., 2003; McClellan et al., 1983; Miller
& Hester, 1986; Rychtarik et al., 2000). These types of clients, when provided a lower level
of care than their condition required, have significantly higher dropout rates and poorer
outcomes (Gastfriend, 2003).
Residential Treatment Placement Under Prop 36
Despite evidence of the benefits of long-term residential treatment for heavy-using clients
and despite Prop 36 funding, the financially strained treatment system has adopted what can
best be described as an “outpatient first” approach to drug treatment. In most counties,
clients must perform poorly in outpatient drug-free treatment before they can be considered
for residential treatment. This is typical of treatment systems across the country (McClellan,
2003). This practice is likely due to the higher cost of residential treatment compared to
outpatient drug-free treatment, limited capacity for residential treatment in many localities,
and the difficulties for capacity expansion due to such issues as zoning and local community
resistance (e.g., NIMBY effects).

174

This increasing reliance on outpatient drug-free treatment, especially under Prop 36 policies
that caused a sharp increase in users referred to treatment, hinders outcomes for heavy using
clients. In the Evaluation of the Substance Abuse and Crime Prevention Act: Final Report
(2006) UCLA showed treatment placements trends under Prop 36 and studied the
relationship between treatment placement and client outcomes.
There was a large increase in the volume of clients presenting for treatment following Prop
36 implementation:
•

•

Many more clients reported the courts or criminal-justice system as their
primary source of referral to treatment after Prop 36 implementation (41% vs.
24%).
The average number of new criminal justice referred treatment admissions
per month for heavy users more than doubled after Prop 36 was
implemented1

The analysis showed significant changes to treatment access and client-composition trends:
•

There was a large increase in the number of heavy users referred to treatment through
the criminal-justice system.

The addiction severity of new clients affects treatment resource requirements. With all
eligible offenders entitled to treatment under the law, the fiscal constraints on the system
meant that most participants were admitted to less expensive treatment programs (usually
outpatient), with limited use of more-expensive options such as residential care.
UCLA studied trends among criminal justice referred clients who were placed into
residential care, for the full population admitted and for the subset of clients considered to be
heavy users. While the absolute number of available residential placements increased
somewhat after Prop 36 implementation, the treatment system was unable to keep pace with
the increase in demand. The percentage of heavy using clients who were accommodated in
residential programs declined significantly following the initiation of Prop 36 (31% of
heavy-users were allocated to long-term residential care before Prop 36, compared with 25%
afterwards).
Among criminal justice referrals, heavy using Prop 36 clients were less likely to receive a
residential placement than non-Prop 36 criminal justice clients with similar addiction
severity. UCLA then studied the characteristics of clients placed into these scarce residential
treatment slots. Young Hispanic males referred to treatment through Prop 36 were
significantly less likely to be placed into residential treatment than similar drug severity
White clients, even after controlling for factors related to treatment placement. Young
Hispanic men referred through Prop 36 were only 66% as likely to receive a residential
placement as similar young White men. However, this treatment placement disparity
observed among Prop 36 offenders diminished for older offenders. There was no meaningful
difference in the likelihood of receiving a residential placement across race/ethnicity for
clients over 35 years. Among Prop 36 offenders, African-American offenders were slightly
1

From an average of 1,280 new episodes per month during the three years before Prop 36 to 2,572 afterwards.

175

less likely to enter residential treatment than White offenders but the difference was not
statistically significant.
To test whether under-treatment across race/ethnicity was of consequence to offender
outcomes, UCLA examined differences in treatment outcomes for Prop 36 offenders. UCLA
compared drug treatment outcomes by modality, as well as differences in criminal justice
outcomes (felony and misdemeanor arrests), for the first year of Prop 36 offenders for 30
months following their entry into Prop 36.
Under Prop 36, treatment completion is a marker of a client’s progress in treatment, but
distinct from the benefits of treatment itself; it is also a marker of a client’s progress towards
meeting the requirements of Prop 36 probation/parole. This is integral to a client’s
successful participation in the Prop 36 program. Among heavy using Prop 36 offenders
UCLA found that, across all of the primary drugs, those who received a residential
placement were significantly more likely to have a successful treatment discharge, and
therefore be in compliance with the terms of their Prop 36 probation, than those who
received outpatient care. Large, statistically significant treatment completion “gaps” were
found for heavy using clients placed into residential care compared with those assigned to an
outpatient program. Forty percent of heavy using Prop 36 offenders admitted to residential
care completed treatment compared to 19% for those who were admitted to non-residential
care (Hawken et al., 2007).
Heavy using Prop 36 clients who were placed into residential care also performed
significantly better on criminal justice outcomes. UCLA compared criminal justice
outcomes for high using Prop 36 clients receiving residential placements compared with
those placed into outpatient care by primary drug after controlling for /ethnicity, age, and
prior arrest history. A residential placement was associated with lower recidivism in the
following 30 months. The effect of treatment placement (residential or outpatient) on
criminal justice outcomes was strongest for Prop 36 offenders reporting methamphetamine
as their primary drug (18% fewer felony, and 17% fewer misdemeanor arrests).
This research showed a number of important disparities in receipt of residential treatment
under Prop 36, which have important implications for setting treatment priorities.
Summary
Disparities in placement into residential care under Prop 36
Heavy using Prop 36 clients are being under-treated compared with non-Prop 36 criminal
justice clients of similar addiction severity. Controlling for client demographics and drug
use patterns, placement rates into residential care were significantly lower for Prop 36 clients
than for non-Prop 36 criminal justice referrals (22% v 31%).
Race/ethnic and gender disparities in treatment placement under Prop 36
Young Hispanic males referred to treatment through Prop 36 are much less likely to be
placed into residential treatment, even after controlling for other factors related to treatment
placement (only 66% as likely to receive a residential placement as similar young White
men). However, there was no meaningful difference in the likelihood of receiving a

176

residential placement across race/ethnicity for clients over 35 years. UCLA found no
placement differences for high severity offenders between Whites and African-Americans
and no placement differences based on gender.
Treatment Provider and Other Key Stakeholder Perceptions of Residential Treatment Under
Prop 36
In the 2007 UCLA Survey of Prop 36 Treatment Providers (see Appendix B), 63% of
providers responded that Prop 36 outcomes would improve if treatment clients had access to
more intensive services, primarily represented by an increase in residential capacity. This
finding was mirrored in the UCLA statewide focus groups with key Prop 36 stakeholders
(see Appendix C). Four dominant themes regarding residential treatment under Prop 36
emerged from the focus groups: the availability of residential slots, the “fail outpatient first”
approach taken in many counties, treatment duration, and other issues related to funding.
Availability of Residential Slots
Access to residential treatment varied substantially across counties. Some county
stakeholders remarked that clients who needed access to residential care could be
accommodated, while others commented that their county was unable to provide the needed
treatment. But the overwhelming majority of focus group participants were of the opinion
that there was insufficient residential care available for Prop 36 treatment clients, in some
cases this was due to limited availability of residential treatment slots, and in some cases due
to limited funds to funds to pay for Prop 36 treatment beds. Certain groups of Prop 36
clients, such as pregnant women, were identified as being especially difficult to place. The
shortage of residential care was regarded as a significant impediment to improved outcomes
under Prop 36.
Common comments from stakeholder focus groups:

“The problem is that there are many contracts bidding for the same facility. Prop
36 is a fee for service, but residential programs have other contracts that they’ve
signed, so there are not enough Prop 36 beds.”
“Our difficulty now is that there aren’t enough female beds, especially perinatal
programs.”
“Waiting lists could be 2 weeks or four months depending on timing. If the wait is
getting long, the county will try not to add to the wait list. Rarely is there an open
bed.”
“[Residential beds have] been completely eliminated from the regular Prop 36
budget; we’re paying for them only through OTP budget. We have hardly any beds
compared to what we used to have. We can expect [our county’s] success rate just
to go right down the toilet.”
Fail Out of Outpatient First
Many focus group participants expressed frustration at being unable to refer Prop 36
treatment clients to an appropriate level of care, given clients’ assessed needs. Many noted
that they were seriously constrained in their ability to place Prop 36 clients into a residential
program. As a result, Prop 36 clients would need demonstrated (often repeated) failures at

177

outpatient treatment before they would be considered eligible for residential care.
Stakeholders expressed dissatisfaction at how this approach was wasting time and resources
and affecting Prop 36 outcomes.
Common comments from stakeholder focus groups:

“If clients started out with residential, outcomes would be better.”
“Need residential as a 1st stop, but we can’t, we must start in Outpatient and have
them fail out of that, which wastes time and resources.”
“Would save money in the long run to start some in residential.”
“More people should be sent to, or assessed to need a residential rather than
outpatient, given the population we’re dealing with. Certainly if your life is in
turmoil with the criminal justice and addiction, you would probably be much better
served in a residential [facility] with more structure than in outpatient.”
“We would save money, in the long run, by putting some of those people in
residential to begin with, because many of them would be in the program for a
shorter amount of time. We, and they, would be more successful.”
Duration in treatment
In the 2007 UCLA Survey of Prop 36 Treatment Providers the median number of days in
treatment reported by residential treatment providers was 90 days. A common theme raised
in the UCLA Focus Groups had to do with time in treatment. Many stakeholder groups
commented that longer residential treatment would improve Prop 36 outcomes.
Common comments from stakeholder focus groups:
“Residential treatment used to be available for 90 days with extensions if needed up
to six months, but now, no matter what a client’s situation is, they have to be out in
90 days.”

“We’ve had to modify the curriculum at our residential modality to accommodate a
ninety day… and even a 30 day stay.”
Funding
Across the board, stakeholders expressed concern about funding cuts to Prop 36. Many
counties relied on the increase in Prop 36 funding in 2006-2007 to finance residential slots.
With Prop 36 funds being cut back in 2007-2008, many stakeholders noted that their
residential treatment would be scaled back from already low levels (a detailed description of
changes to Prop 36 funding is provided later in this chapter).
Residential Treatment Findings from 2005 Stakeholder Study
In 2005, Gelber and colleagues conducted a study of Prop 36 stakeholders. This study
identified funding as the primary concern among Prop 36 stakeholders. In addition to
expressed concerns regarding the limited availability of residential treatment due to funding
constraints, stakeholders noted two additional resource limitations which purportedly
affected Prop 36 participant outcomes: access to continuity of care services and the
availability of sober-living environments. This is of particular concern in counties where
funding limitations resulted in reductions in duration in treatment, and the amount of after-

178

care provided under Prop 36. Given the (often) long lag between treatment completion and
completion of Prop 36 probation, this leaves many Prop 36 clients vulnerable to relapse.
Common comments from stakeholder focus groups:

“It is a money issue. The majority of Prop 36 clients would be better served in
residential rather than outpatient.”
“Due to severe budget cuts, as of July 1, 2007, residential slots are only available via
the OTP contract, and are no longer paid for through Prop 36. There are hardly any
beds and clients get one chance at residential. If they fail, they cannot go back. The
county’s success rate is expected to sharply decline. Many of the residential
programs have spent their money before the end of the fiscal year and now have a
shortage of beds.”
“Due to finances, the county is becoming more “cookie-cutter,” giving everyone
the same treatment.”
“(In our county)… there is capacity, just no funding.”
“As outpatient providers, we’re getting clients that cannot be serviced at our level
of treatment. They need much more higher levels of treatment, but because a set
amount of dollars need to be spent for so many clients, we get a lot of clients that,
you know, just, we just cannot – we can do the best that we can. But under the
circumstances, we are setting them up for failure, because they truly do need
residential treatment.”
“I just found out one of, one of my big player residentials is going to be out of
money in about two and a half months. The need was so great. We’ll be lucky if
OTP money lasts ‘til October.”
“More treatment in a residential modality, you know, increases the potential for
success rates. But, because of the funding cuts, et cetera, et cetera, et cetera…”
“I just don’t have funding to buy residential beds anymore.”
County plans as reflected in OTP applications
Many of the OTP applications submitted to ADP reflect county concerns regarding the
delivery of residential treatment services to Prop 36 treatment clients. Thirty-nine counties
submitted applications for OTP funding, which detailed how OTP funds would be used.2
UCLA coded county responses to identify common funding requests. The primary OTP
request (across all utilization strategies) was for funding to increase the use of residential
services. Common concerns identified were the lack of residential treatment slots (either due
to limited availability of slots, or due to limited funding to pay for slots), long delays
between assessment and entry into treatment, and the insufficient duration of residential
treatment offered.

Figure 9.1 shows three residential strategies listed in OTP applications. Fifty-nine percent of
counties planned to use OTP funds to increase the utilization of residential treatment services
2

Technical assistance was available for the counties in preparing the applications. Fewer than 70% of the
counties applied for OTP funds. Unused funds were returned to the General Fund.

179

and 49% of the counties planned to use OTP funds to expand residential bed capacity.
Thirteen percent of the counties planned to use OTP funds to extend the length of residential
treatment stays.
Fifty-nine percent of the counties planned to use OTP funds to help reduce treatment delays.
The use of OTP funds to reduce treatment delays was not specific to residential care; the
listed strategies to reduce treatment delays applied to residential and outpatient treatment.
Figure 9.1
Percentage of County Listed Strategies Related to Residential
Care in OTP Applications
70
60

59%
49%

Percent

50
40
30
20

13%

10
0
Increase utilization of Expand residential bed
residential treatment
capacity
services

Extend residential
treatment stays

Note: Data are from the County applications for OTP funding submitted to the California
Department of Alcohol and Drug Programs. Thirty-nine counties submitted requests for
OTP funds. Data were coded for common funding requests.

Barriers to Expanding Access to Residential Treatment
Prop 36 funding
Data from UCLA’s statewide stakeholder focus groups and surveys show that funding was
the primary barrier to expanding access to residential treatment. The final report of the first
round of the Prop 36 evaluation included a study of adequate funding under Prop 36. This
analysis showed that Prop 36 was under-funded and UCLA made a recommendation to fund
the program at a level of $229 million in 2006 dollars. Table 9.1 shows the funding
allocations for Prop 36 from fiscal year 2001-02 through 2007-08.

Table 9.1: Prop 36 Funding Allocations 2001 to 2008 (in $ millions)

SATTFa
OTPb
Total

Fiscal year ‘01-‘02
through ‘05-‘06
$120
$120

2006-2007

2007-2008

$120
$25
$145

$100
$20
$120

Note: a The Prop 36 Substance Abuse Treatment Trust Fund
b
The Offender Treatment Program

180

Difference ‘06‘07 to ‘07-‘08
($20)
($5)
($25)

The funding mechanism originally written into the law was an annual allocation of $120
million (with no provision to adjust for inflation) to counties for five years, ending in June
2006. Thereafter, funding decisions reverted to the administration and the legislature.
The 2006-07 state budget provided for $145 million from the General Fund, which included
$25 million for OTP. The OTP required a 10% funding match from counties. Governor
Schwarzenegger cut 2007-08 funding for Prop 36; to a total of $100 million, plus $20 million
to be distributed through the OTP. In terms of real purchasing power (i.e., accounting for
inflation erosion), the $120 million allocated for Prop 36 and OTP in 2007-08 represents a
$17 million effective decrease in program spending, compared to the initial $120 million
allocation when the program was implemented in 2001.3 Using the medical price index (a
closer match of composite goods and services), the effective funding cut for 2007-08 from
the 2001-2002 level is $27 million. The effective funding cut for 2007-08 from the 2006-07
level is $31 million.4
Counties and treatment providers are vulnerable to funding decisions made at the state level.
UCLA’s 2007 Survey of Prop 36 Treatment Providers shows treatment providers are
extremely limited in their ability to secure supplemental funding.
In responding to the question: “Has your program been able to secure supplemental funding
(other than Prop 36 trust funds, SATTA5, county general funds, and fees collected from Prop
36 clients) to facilitate the operation of Prop 36 (e.g., grants from private or federal agencies,
or the Offender Treatment Program)?” only 32% of Prop 36 treatment providers reported that
they were able to secure supplemental funding to facilitate operations. This included access
to OTP funds, which have since been reduced.6
The Prop 36 funding cuts will pose a significant challenge to residential treatment provision
in the face of inflation erosion and budget pressures, counties are likely to cut back, rather
than expand, on existing treatment services.
Other Barriers

Funding is not the only constraint limiting the expansion of residential services. New
treatment providers often face NIMBY issues (Not in My Back Yard) when nearby residents
resist the opening of the proposed treatment site. Zoning laws and community resistance,
together, pose a substantial barrier to capacity expansion.
Focus Group participants noted that there were often mismatches between the residential
treatment slots that were available in a county, and slots that were needed. In particular,
participants noted that there were insufficient treatment slots for women, and a shortage of
perinatal programs. This highlights the need for routine Needs Assessments to keep counties
3

Using the consumer price index (CPI) from the Bureau of Labor Statistics.
Medical costs increased by 4.2% from 2006 to 2007. To maintain purchasing power, the 2007-08 Prop 36
budget would have had to increase to $151 million. The actual funding allocation of $120 million represents a
$31 million real spending cut.
5
Substance Abuse Treatment Accountability
6
Respondents were not asked to specify the sources of alternative funding available.
4

181

updated about the profile of their Prop 36 treatment caseloads and the treatment needs of
their clients. This will help guide treatment expansion plans to ensure that treatment is added
where it is needed the most.
Recommendations
A number of recommendations follow from our examination of statewide data:
Improved treatment matching
For heavy using clients, offender treatment and criminal justice outcomes were better if the
offender was placed in residential care. Resources should be allocated to ensure suitable
treatment matching to offenders’ needs. This may require capacity expansion, locating
treatment centers near areas of high need, and greater use of residential services. This will
require routine needs assessments.
Prioritize residential care for young Hispanic males
Young Hispanic males are currently under-served. Expanding residential treatment to young
Hispanic males should be prioritized. This may require capacity expansion in locations
conveniently located to young Hispanic offenders.
Prioritize care for offenders reporting methamphetamine as their primary drug
Treatment placement (residential or outpatient care) mattered most for clients who entered
Prop 36 with methamphetamine as their primary drug. It would be cost-effective to
prioritize methamphetamine users for residential care.
Increase use of sober-living environments and continuing care services
For many Prop 36 clients, there is a significant delay between treatment completion and
completion of Prop 36 probation. Providing sober-living options and continuing-care
services will reduce the likelihood that a client will relapse between treatment completion
and the completion of the terms of their Prop 36 probation. This is particularly important in
counties that have responded to funding constraints by reducing the use of residential care
and the required duration in residential treatment.

References
Gastfriend, D.R. (2003). Addiction Treatment Matching: Research Foundations of the
American Society of Addiction Medicine (ASAM) Criteria. New York: Hawthorne
Press.
Gastfriend, D.R., & McClellan, A.T. (1997). Treatment matching: Theoretical basis and
practical implications. Alcohol and Other Substance Abuse, 81, 945-965.
Hawken, A, Anglin, D.M., & Conner, B.T. (2007). Treatment Differences. In
Evaluation of the Substance Abuse and Crime Prevention Act: Final Report.
Magura, S., Staines, G. Kosanke, N., Rosenblum, A., Foote, J., DeLuca, A., & Bali, P.
(2003). Predictive validity of the ASAM Patient Placement Criteria for naturalistic
matched versus mismatched alcoholism patients. American Journal on Addictions,
12, 386-397.

182

McClellan, A.T., Woody, G.E., Luborsky, L., O'Brien, C.P. & Druley, K.A. (1983).
Increased effectiveness of substance abuse treatment: A prospective study of patienttreatment ‘matching’. Journal of Nervous and Mental Disease, 171, 597-605.
McClellan, A.T., Carise, D., & Kleber, H.D. (2003). The national addiction treatment
infrastructure: Can it support the public's demand for quality care? In. J. Jaffe & D.
Rosenbloom (Eds.), Contemporary Issues in Addiction Treatment. Binghamton, NY:
Haworth Press.
Miller, W.R., & Hester, R.K. (1986). The effectiveness of alcoholism treatment: What
research reveals. In Miller, W.R. & N. Heather (Eds.), Treating Addictive Behaviors:
Processes of Change. New York: Plenum Press.
Rychtarik, R.G., et al. (2000). Treatment settings for persons with alcoholism: evidence
for matching clients to inpatient versus outpatient care. Journal of Consulting and
Clinical Psychology, 68, 277-289.
Sharon, E., Krebs, C., Turner, W., Desai, N., Penk, W., & Gastfriend, D. R. (2003).
Predictive validity of the ASAM patient placement criteria for hospital utilization.
Journal of Addictive Diseases, 22, 61-76.

183

184

Chapter 10: Testing and Sanctions for Proposition 36 Probation
Violations
Angela Hawken, Ph.D. and Annie Poe, M.P.P.

This chapter considers the role for sanctions in response to non-compliance with the
terms of Prop 36. In an earlier Prop 36 evaluation report, UCLA recommended a greater
use of drug testing information to determine additional services or intermediate sanctions
that are enhanced with each successive violation.
Applications for OTP funds reflected county interest in strengthening criminal justice
supervision under Prop 36. Many counties submitted plans to increase the number of
probation staff and to expand drug testing. Enhanced community supervision was listed
among the top three priorities for OTP spending.
The basic tenets of a graduated sanctions program (swift, certain, and parsimonious use of
sanctions) have strong theoretical underpinnings and are well supported in the literature.
There are many sanctions options available, including spending days in a jury box,
intensifying treatment, community service, house arrest, and more intensive probation
supervision. Moreover, there is a small but growing evidence base on testing and jail
sanctions programs that shows that swift and certain, but modest, jail sanctions can bring
about positive behavior change. These programs improved outcomes only when
probation conditions and consequences were clearly articulated to probationers, and when
violations were dealt with consistently and with certainty. Where consistency was
lacking, testing and jail sanction programs have failed.
Expanding the conditions of Prop 36 probation to include sanctions for non-compliance
has been controversial, especially in regard to applying brief jail stays (also known as
“flash incarceration”). Senate Bill 1137 was passed by the legislature in 2006 and
provided discretion to judges to give short jail stays of up to ten days to motivate
treatment and probation compliance. This bill was opposed in court on the grounds that
jail sanctions would violate the intent of California voters who passed the Prop 36
initiative, and an injunction was issued. But among key stakeholders involved in
managing Prop 36 probationers there has been growing support for sanctions options
within Prop 36. In UCLA’s statewide focus groups, public defenders were the only
stakeholder group that did not recommend enhanced sanctions (including flash
incarceration) as a condition of Prop 36 probation. A survey of treatment providers
indicated that 80% supported flash incarceration for continued non-compliance as a
mechanism to improve treatment outcomes.
There are three key barriers to the use of sanctions under Prop 36: 1) legal barriers due to
the original language of the Prop 36 initiative and the court injunction of Senate Bill 1137
limit the types of sanctions allowable under Prop 36, 2) solving the public management
problem of ensuring that all the key players in the Prop 36 system implement a system of
graduated sanctions consistently, and 3) designing process changes to reduce the work
burden that would result.
185

Testing and Sanctions Programs
There are many sanctions options available, including spending days in a jury box,
intensifying treatment, house arrest, and more intensive probation supervision. In this
chapter UCLA focuses on testing and sanctions programs because these programs were
recommended by Prop 36 stakeholders most frequently.
Testing and sanctions programs require frequent drug testing and impose consistent
sanctions for violations. A graduated sanctions package includes the use of sanctions that
increase in intensity for successive violations. The use of testing and sanctions as a
mechanism to motivate compliance with treatment and other terms of Prop 36 probation has
been controversial, particularly with regard to the use of short jail stays, or “flash
incarceration”. Prop 36 does not provide for drug testing and incarceration as a sanction for
noncompliance. Within a year of the implementation of Prop 36, Senate Bill 223 was passed
by the legislature, which allowed separate funding for testing, on the conditions that drug
testing under Prop 36 be used as a treatment tool and that a failed drug test alone would not
constitute grounds for drug-related probation violations (PC section 1210.5). As written,
Prop 36 probationers may only be sentenced to a jail term if they have accumulated two prior
violations.
In 2006, citing less than optimal Prop 36 treatment entry and treatment completion rates and
limited probationer accountability as a key concern, Senator Ducheny introduced Senate Bill
1137. This bill allowed discretion to judges to impose short jail sanctions for violations of
the terms of Prop 36 probation.
This bill would authorize a court to also order incarceration for a specified
period, in order to enhance treatment compliance, and in some
circumstances, to order the defendant to enter a residential drug treatment
program, if available, or be placed in a county jail for not more than 10
days for detoxification purposes only.
(SB 1137)

Senate Bill 1137 was strongly opposed by advocacy groups on the grounds that such
sanctions violated the original intent of the voters. The matter was taken to court and an
injunction followed. In its current form, Prop 36 prohibits incarceration as a condition of
probation.
Many Prop 36 stakeholders have noted that the lack of sanctioning options under Prop 36 has
resulted in high levels of offender non-compliance with the terms of their Prop 36 probation
and in low levels of participant motivation (stakeholder perceptions on sanctions under Prop
36 are described below).
Testing and Sanctions in Theory
Here UCLA summarizes the theoretical underpinnings of testing and sanctions and review
empirical evidence on testing and sanctions in practice. Sanctions here refer to a penalty
imposed for non compliance. Testing and sanctions programming has a strong theoretical
basis. When applied swiftly and consistently, testing-and-sanctions can promote behavioral
change. Testing and sanctions programs that follow these basic tenets (clearly articulated

186

sanctions applied in a manner that is certain, swift, consistent, and parsimonious) are
research based:
A clearly defined behavioral contract
Probationers should be informed about the conditions for compliance with the terms of their
probation and consequences for each violation should be carefully explained (Taxman,
1999). A clearly defined behavioral contract has been shown to enhance perceptions of the
certainty of punishment which improves compliance (Grasmack & Bryjak, 1980;
Paternoster, 1989; Nichols and Ross, 1990; Taxman, 1999).
Consistency
All agents in the criminal justice system and treatment providers need to enforce the stated
rules (Harrell and Smith, 1996). The consistent application of a behavioral contract has been
shown to improve compliance (Paternoster et al., 1997) and enhance perceptions of fairness
(Taxman, 1999).
Swift delivery
Sanctions should be delivered in a timely fashion (Taxman, 1999). A swift response to
infractions improves the perception that the sanction is fair (Rhine, 1993). The immediacy,
or celerity, of a sanction is also vital for shaping behavior (Farabee, 2005).
Parsimony
Parsimonious use of punishment (i.e., the least amount of punishment necessary to bring
about the desired behavior change) enhances the legitimacy of the sanction package and
reduces the potential negative impacts of tougher sentences, such as long jail or prison stays
(Tonry, 1996).
Awareness of dignity (also called “procedural justice”)
Maintaining an appreciation for probationer’s dignity through the process of behavior change
is also important (Taxman, 1999). The supervision process itself has an independent effect
on compliance (Taxman, 1999). The manner in which sanctions are imposed and enforced
by judges, probation agents, and other actors in the criminal justice system shapes the
probationer’s views regarding the legitimacy of these authority figures and the sanctions
imposed, and affects the probationers’ decision to comply with the rules (Tyler, 1990;
Paternoster et al., 1997). Fair and respectful management of probationers enhances
compliance (Bazemore & Maloney, 1994; Braithwaite, 1989; Sherman, 1993).

Despite strong theoretical underpinnings, there have been relatively few instances of
widespread testing and sanctions programs implemented in practice.
Testing and Sanctions in Practice
Some states and local areas nationally are using testing and sanctions programs both to
improve entry into treatment and to keep clients in treatment. The drug-testing with
sanctions programs that have been implemented in various jurisdictions show degrees of
success that positively correlate with how reliably the conditions of probation are enforced
(Kleiman, 2001, Harrell & Roman, 2001).

187

Here UCLA reviews five testing-and-sanctions programs:
•
•
•
•
•

The Washington D.C. Superior Court experiment
Maryland’s Break the Cycle
New York’s Drug Treatment Alternative-to-Prison program (DTAP)
Hawaii’s HOPE probation
Georgia’s graduated sanctions program for parolees

The Washington D.C. Drug Court Experiment
In 1993, drug offenders in Washington, D.C. became part of a trial evaluation of treatment
and sanctions. This experiment was in reaction to research that showed the direct link
between drug use and crime. Lengthy sentences against drug users did not dissuade or
prevent these offenders from committing more crimes. The Washington, D.C. Superior Court
decided to try a different approach by using treatment and sanctions as a means of deterrence
(Harrell et al., 2001).

The experiment consisted of a randomized controlled trial (RCT). The majority of
participating drug offenders were male and in their early thirties (similar to the profile of
Prop 36 probationers). Study participants were randomly assigned to one of three dockets:
Docket 1 (The “standard” Docket): these offenders received the normal process of drug
testing and judicial monitoring with no sanctions for failed drug tests.
Docket 2 (The Treatment Docket): these offenders were assigned to intensive treatment.
Docket 3 (The Sanctions Docket): concentrated on immediate sanctions (the graduated
sanctions package began with three days in a jury box, then to three days in jail, then 5-7
days in detoxification, then 7 days in jail) for failed urinalyses or missed appointments, with
treatment provided if needed or desired.
The results of the RCT indicated that sanctions combined with voluntary treatment were the
most effective form of deterrence. Offenders assigned to treatment without the threat of
sanctions had fewer positive urinalyses compared with the standard docket, but no difference
in follow-up recidivism. Offenders assigned to the sanctions program, had lower drug use
and lower follow-up recidivism compared with the standard docket (follow-up recidivism
was lower for the sanctions group than the treatment group). The findings from this study
indicate that the use of testing and sanctions was effective in reducing drug use and
recidivism.
Maryland’s Break the Cycle
Maryland’s Break the Cycle (BTC) was implemented as part of Maryland’s effort to improve
the benefits of community supervision and treatment for drug-involved offenders. The BTC
program combined drug testing, treatment, and sanctions to reduce criminal behavior and
drug use. The BTC program was developed to create a system without boundaries, where
criminal justice officials and treatment providers worked together and shared information to
improve offender outcomes. But Maryland did not manage to bring about the collaboration
required to make the program function well. The implementation of the program improved

188

over the first three years in a number of key areas (the sanctions rate per positive test rose
from 3% to 56%, and the sanctions rate for no-shows increased from 1% to 65%), but
warrants and revocations continued to be slow (taking an average of 137 days in year 3) and
sanctions were not applied consistently (Taxman et al., 2002). In many cases, a positive
drug test led to a verbal or written warning, and in other cases, there was no sanction at all.
Testing positive for drug use rarely led to an arrest warrant. These management and
collaboration problems undermined the potential effectiveness of the program. The
implication of the BTC study is that implementing a solid testing and sanctions program is
difficult, and when sanctions are not applied consistently, will likely fail.
New York’s Drug Treatment Alternative-to-Prison program (DTAP)
The District Attorney of Kings County, New York implemented the Drug Treatment
Alternative-to-Prison program to ease prison overcrowding and reduce recidivism. The
program incorporated residential treatment and sanctions for failure to complete treatment.
If the offender absconds or fails to complete the treatment program, the District Attorney’s
special warrant enforcement team is immediately dispatched and the offender is brought to
court and incarcerated for their outstanding charges. In the evaluation of the program, DTAP
offenders were compared with a matched comparison group of offenders who were
processed in the regular criminal justice program. By October 2007, 2,500 offenders had
participated in the program (Kings County District Attorney, 2007). Retention rates were
high (76%) (Kings County District Attorney, 2007). Those who failed to complete the
conditions of the program were rapidly returned to court for sentencing. Across all program
participants (completers and non-completers), DTAP participants had a 26% lower re-arrest
rate and were 67% less likely to return to prison than offenders in the comparison group (The
National Center on Addiction and Substance Abuse, 2003). Certainty and swiftness of
punishment for failure was a key component in DTAP’s success.
Hawaii’s HOPE Probation
Hawaii’s HOPE Probation (Hawaii’s Opportunity Probation with Enforcement) provides the
most recent evidence of outcomes under a strictly enforced testing-and-sanctions program.
The program provides for close monitoring of probationer behavior, and rapidly punishes
violations (including positive drug tests) with mild sanctions -- typically a few days in jail,
with the number of days served increasing for successive violations.

The key features of the HOPE testing-and-sanctions package include:
•
•
•

•

•

Random testing at least once a week.
Modest jail sanctions in response to positive drug tests, and no-shows: Typically a
few days for a first violation, with sentence length increasing gradually.
A formal warning regarding the terms of the testing and sanctions program to the
probationer in open court, putting him or her on notice that all probation violations
will have immediate consequences.
As short a time as possible between violations and sanctions, typically within 24
hours but never more than 72 hours. For offenders with paycheck jobs, the first
sanction is often deferred to the weekend.
Quick service of bench warrants on those who abscond.

189

•

Enhanced treatment services for those who prove unable to refrain from drug use on
their own.

The pilot study of HOPE probation was started in 2004. Probation officers were asked to
identify the drug-involved probationers on their caseloads who had demonstrated repeated
non-compliance with the terms of their probation, including multiple positive drug tests, and
posed a high risk of revocation. The list of probationers identified were rank-ordered by
risk. The top half of the high-risk probationers (those identified to be of highest risk) were
placed into the HOPE program, the bottom half of the high-risk probationers identified were
used as the comparison group. The number of dirty urinalyses among HOPE probationers
fell by 85% over baseline by 3 month follow-up (compared with a 35% increase for the
comparison group). Both positive urinalyses and missed appointments fell precipitously as
exposure to HOPE increased (compared with the comparison group, among whom the
percentage of probationers with positive drug tests and missed appointments increased as
time on probation increased). Due to the success of the HOPE program, the pilot program
(which began with a single judge) has now been adopted by all nine circuit court judges in
the state, and the legislature has appropriated funds to expand the pilot program to include
one fourth of all the state’s felons on probation. The expanded program is yielding success
rates similar to the pilot. A randomized controlled trial of the HOPE program is currently
underway.
Georgia’s graduated sanctions program for parolees
In 2005 the Reentry Policy Council (convened by the Council of State Governments) issued
a report detailing the state of the art in community supervision. Georgia’s graduated
sanctions package was highlighted in the report (Reentry Policy Council, 2005). Georgia
first implemented a graduated sanctions package for parolees in 1991; by 1998 the state had
a fully implemented graduated sanction system in place. Between 1998 and 2002 the state
had a 12% increase in successful parole completion (Reentry Policy Council, 2005). In 2003
the state of Georgia amended their graduated sanctions program to incorporate positive
adjustments for compliance and prosocial behaviors, such as negative urinalyses and stable
employment (La Vigne & Mamalian, 2004). These ammendments led to the development of
a guide that describes their graduated sanctions package called the Behavior Response and
Adjustment Guide, also referred to as the BRAG (La Vigne & Mamalian, 2004). Behaviors
(both positive and negative) are graded on a continuum (low, medium, and high) and
responses to behaviors are clearly detailed.

Together, these programs illustrate the potential for testing and graduated sanctions programs
to improve offender outcomes. A key to the success of these programs is consistency and
certainty. Conditions of probation need to be clearly articulated to probationers, and each
violation of probation consistently penalized. Where probation terms are not consistently
enforced, these programs fail to deliver on their promise.
Prop 36 Treatment Provider Perceptions on Why Prop 36 Clients Did Not Complete
Their Planned Treatment
In an earlier chapter, UCLA reported 32.2% of Prop 36 participants completed the drug
treatment program to which they were mandated. The UCLA 2007 provider survey included
questions to determine treatment providers’ perceptions of why Prop 36 clients did not
190

complete their treatment program (see Figure 10.1). Seventy-four percent of treatment
providers responded that many Prop 36 clients did not complete treatment because they were
unwilling to comply with the terms of Prop 36 requirements, and 63% responded that Prop
36 clients did not complete treatment because they lacked motivation.1 Smaller percentages
of providers responded that non-compliance was due to transportation problems (19%),
conflicting work schedules (17%), a lack of stable housing (30%), and family responsibilities
(18%). At least from the perspective of treatment providers, strategies to achieve greater
compliance (i.e., completion of treatment) would need to be primarily targeted at clients’
unwillingness to comply with Prop 36 requirements and at their low motivation. Further
research is required to identify the causes of low motivation of Prop 36 treatment clients. If
low motivation is a result of limited probationer accountability under Prop 36, then a
graduated sanctions package may be an appropriate response. If low motivation is a result of
inappropriate treatment matching, insufficient treatment intensity, or insufficient treatment
duration, appropriate service delivery should be prioritized.
Figure 10.1
Treatment Provider Perceptions of Why Prop 36. Clients Did Not Complete
Their Planned Treatment.
80

74%

70

63%

Percent

60
50
40
30
20

30%
19%

17%

Transportation

Work schedule

18%

10
0
Housing

Family
responsibilities

Unwilling to
comply

Motivation

Note: Data are from the UCLA 2007 Prop 36 Treatment Provider Survey. See Appendix C for a
description of the survey. Prop 36 treatment providers were asked: “To what extent do the following
(list shown in Figure 10.1) describe reasons why Prop 36 clients have not completed their planned
treatment duration at this program.” Respondents were asked to rate the reasons on a four-point scale
(Not at all, Limited Extent, Moderate Extent, Great Extent). The bars in Figure 10.1 represent the
percentage of treatment providers who responded “Moderate Extent” or “Great Extent”.

How Prop 36 Treatment Providers Respond to Positive Urinalyses
The UCLA 2007 Prop 36 Treatment Provider Survey included questions on how treatment
providers responded to positive urinalyses. Figures 10.2 – 10.6 show the extent to which
providers take specific actions to a positive drug test.2 Seventy-six percent of providers that
completed the survey reported that they commonly adjusted a client’s treatment plan in
response to a positive drug test. For 66% of providers, the response was a change in the
level of care. Seventy percent of providers increased the frequency of drug testing. For 44%
1

Respondents were not asked to define motivation or provide details of possible mechanisms underlying their
clients’ lack of motivation.
2
The survey did not distinguish between responses to a single test or to multiple tests over time.

191

of providers, a positive drug test would commonly result in discharge from the program with
a referral to another program, and 9% would typically be discharged without a referral.

Percent

Figure 10.2
An adjustment is made to client’s treatment plan
60
50
40
30
20
10
0

54%

22%
16%
8%

Not At All

Limited
Extent

Moderate
Extent

Great Extent

Note: Data are from the UCLA 2007 Prop 36 Treatment Provider Survey. The results reflect
responses from 87 randomly selected Prop 36 treatment providers to the question: “To what
extent is an adjustment made to the client’s treatment plan if the Prop 36 client tests positive
for drugs at your program?”

Percent

Figure 10. 3
A change is made to client’s level of care
40
35
30
25
20
15
10
5
0

34%

32%

Moderate
Extent

Great Extent

20%
14%

Not At All

Limited
Extent

Note: Data are from the UCLA 2007 Prop 36 Treatment Provider Survey. The results reflect
responses from 87 randomly selected Prop 36 treatment providers to the question: “To what
extent is a change made to the client’s level of care if the Prop 36 client tests positive for
drugs at your program?”

192

Percent

Figure 10.4
Frequency of drug testing is increased
40
35
30
25
20
15
10
5
0

38%
32%
17%

Not At All

13%

Limited
Extent

Moderate
Extent

Great Extent

Note: Data are from the UCLA 2007 Prop 36 Treatment Provider Survey. The
results reflect responses from 87 randomly selected Prop 36 treatment providers to
the question: “To what extent is the frequency of drug testing increased if the Prop
36 client tests positive for drugs at your program?”

Percent

Figure 10.5
Discharged with a referral to another program
40
35
30
25
20
15
10
5
0

34%
22%

Not At All

20%

Limited
Extent

Moderate
Extent

24%

Great Extent

Note: Data are from the UCLA 2007 Prop 36 Treatment Provider Survey. The
results reflect responses from 87 randomly selected Prop 36 treatment providers to
the question: “To what extent is a client discharged with a referral to another
program if the Prop 36 client tests positive for drugs at your program?”

193

Percent

Figure 10.6
Discharged without a referral to another program
80
70
60
50
40
30
20
10
0

72%

19%

Not At All

Limited
Extent

5%

4%

Moderate
Extent

Great Extent

Note: Data are from the UCLA 2007 Prop 36 Treatment Provider Survey. The
results reflect responses from 87 randomly selected Prop 36 treatment providers to
the question: “To what extent is a client discharged without a referral to another
program if the Prop 36 client tests positive for drugs at your program?”

Treatment Provider and Other Key Stakeholder Perceptions of Testing and Sanctions
A common misconception is that the criminal justice system is the driving force behind the
sanctions debate and that treatment providers oppose the use of sanctions to motivate
treatment compliance. UCLA surveys of Prop 36 treatment providers and focus groups with
key-stakeholders suggest otherwise.
Results from the UCLA Treatment Provider Survey

Statewide surveys of Prop 36 treatment providers show growing support for the use of a
testing and graduated sanctions program (including brief jail stays) as a means to motivate
treatment entry and treatment compliance. The 2005 UCLA Treatment System Impact
Program Survey showed that over half of the Prop 36 treatment providers were in favor of
strengthening sanctions under Prop 36 (this survey did not specify the use of jail sanctions).
Two years later, the 2007 UCLA Prop 36 Treatment Provider Survey included a question
specifically related to jail sanctions for non-compliance. The percentage of treatment
providers who supported expanded sanctions was high: 80% of treatment providers
recommending brief jail stays for continued non-compliance to improve treatment outcomes.
Figure 10.7 summarizes providers’ responses to the question “Do you think treatment
completion at your program would be improved if Prop 36 clients were given brief jail stays
for continued treatment noncompliance?” Of 87 providers, 80% responded that jail
sanctions would improve treatment compliance, 19% were of the opinion that jail sanctions
would not promote treatment compliance, and 1% responded that jail sanctions might
improve compliance.

194

Figure 10.7
Providers’ Perceptions – would jail sanctions improve treatment completion?

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

80%

19%
1%
No

Maybe

Yes

Note: Data are from the UCLA 2007 Prop 36 Treatment Provider Survey. The results reflect
responses from 87 randomly selected Prop 36 Treatment Providers to the question: “Do you think
treatment completion at your program would be improved if Prop 36 clients were given brief jail stays
for continued treatment noncompliance?”

The Treatment Provider Survey included open-ended responses. Many providers reinforced
their position on sanctions under Prop 36 in open-ended responses that supported the notion
of a graduated sanctions package under Prop 36. The dominant themes that emerged from
open-ended responses were:
•
•
•

There should be greater offender accountability under Prop 36,
There should be stronger consequences for non-compliance and relapse
under Prop 36,
Clients should be given short jail stays for continued non-compliance.

Providers commented that many clients did not take their treatment seriously. Clients
recycled through treatment many times and believed that they could do so without
consequences. Providers noted that the delayed consequences for probation violations by the
criminal justice system required under Prop 36 created the impression that the Prop 36
program is not serious and that there are no serious consequences. Providers also
recommended that issues regarding treatment and probationer accountability be addressed
state-wide, with greater consistency across counties.
Graduated sanctions, including brief jail stays, compared favorably against other methods to
motivate treatment compliance in cases of continued non-compliance. Figure 10.8 shows
how treatment providers rated brief jail stays as a means to improve treatment compliance,
195

compared with two other options: (1) the use of treatment reminder calls regarding treatment
admission and participation, and (2) more-intensive treatment.
Of the three options posed, brief jail stays was rated as likely to improve outcomes by the
highest percentage of treatment providers.
Fig 10.8
Prop 36 Treatment Providers’ Perceptions of Practices that would Improve
Treatment Compliance

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

80%
63%

37%

Treatment
More-intensive Brief jail stays
reminder phone
treatment
for continued
calls
non-compliance
Note: Data are from the UCLA 2007 Prop 36 Treatment Provider Survey. The results reflect responses
“Yes”/ “No” responses from 87 randomly selected Prop 36 Treatment Providers to the question: “Do
you think treatment completion at your program would be improved if Prop 36 clients were given brief
jail stays for continued treatment noncompliance?” The numbers reported reflect the percentage of
providers who responded “Yes”.

Results from UCLA Stakeholder Focus Groups
To solicit perspectives on testing and graduated sanctions programs, the issue of sanctions
was raised in focus groups conducted with key stakeholders involved in managing Prop 36
participants (see Appendix A for a description of focus group methodology, including detail
on participant characteristics). These included focus groups with judges, district attorneys,
public defenders, police, probation officers, county administrators, and drug treatment
providers. There was widespread support for graduated sanctions (including brief jail stays
for continued non-compliance) under Prop 36 across nearly all stakeholders to improve

196

accountability and treatment motivation. Public Defenders were the only key stakeholder
group that did not recommend introducing testing and sanctions as a condition of Prop 36.3
Current County Plans for Testing and Graduated Sanctions as Reflected in OTP
Applications
The Offender Treatment Program (OTP) provided financial assistance to counties to enhance
services for Prop 36 offenders, including community supervision. Many of the thirty-nine
counties that submitted OTP applications committed to enhancing criminal justice
supervision under Prop 36, and increasing the use of drug testing. Expanding criminal
justice supervision was the third most common goal listed in OTP applications (following
closely behind expanding residential treatment and reducing treatment delays).
Probation support
Twenty counties submitted plans to increase probation supervision capacity. Fifteen of the
thirty-nine counties applying for OTP funds planned to add probation officers, four planned
to add deputy probation officers, and two counties planned to use OTP funds to hire
probation assistants.
Supporting Drug Court Models
Nineteen of the thirty-nine counties applying for OTP funds reported that they use a drug
court model.4 Six counties reported they would use OTP funds to adopt a drug court
approach.
Drug Testing
Eleven counties reported that they planned to use OTP funds to greater utilize probation
and program drug test results. However, currently OTP funds may not be used for drug
testing.
Challenges to Implementing a Testing and Graduated Sanctions Program
Prop 36 offenders’ rates of compliance with the terms of their Prop 36 probation have been
of concern since the law’s implementation. Close to one third never appear for treatment and
only a third of those who do enter treatment will complete successfully. Many key
stakeholder groups (including treatment providers and criminal justice officials) are
recommending graduated sanctions, up to and including flash incarceration, to motivate
compliance with program terms and to improve treatment outcomes. Use of incarceration as
a sanction may require a court ruling on Senate Bill 1137 or successful passage of a new
proposition amending the Prop 36 law to authorize its use. Apart from legal barriers, there
are many additional potential impediments to a well functioning testing and sanctions
component.

The evidence on graduated sanctions suggests that consistency in providing sanctions is key.
The crucial difficulty in implementing a successful program is ensuring cooperation among
the key players in the system. To be effective, a sanctions program requires that judges,
probation officers, police, corrections officials, and treatment providers all work together.
3

No elaboration was provided for the reasons underlying the Public Defenders’ resistance to the use of testing
and sanctions.
4
The counties did not detail the components of a drug court model that were being used.

197

But wherever a testing and sanctions program has been implemented, and sanctions were
actually delivered adhering to best practices, impressive reductions in drug use and
improvements in probation compliance have been observed. Where the sanctions weren’t
delivered, the expected potential was not realized.
Recommendations for Implementing a Testing and Graduated Sanctions Program
There are a limited number of instances where testing and graduated sanctions programs
including flash incarceration have been implemented, but it has never been tried on a Prop 36
population. UCLA’s recommendation would be to begin with a small pilot project in one
county or a few counties that volunteer to be included. Stakeholder input should be solicited
to determine the terms of the testing and graduated sanctions programs for each stage of the
sanctions process, and to determine what authority would lie with each agency. Developing
clear and specific protocols for cooperation among key stakeholders is essential to a
successful testing and sanctions program. The pilot study could be used to identify process
improvements to reduce the work burden that would fall on many key players in the Prop 36
system, in particular, probation officers and court staff. Incentives should be provided to
probationers who comply with the terms of their probation (this is the case with Hawaii’s
HOPE model and the Georgia graduated sanctions package for parolees ─ the BRAG
model). Such incentives may include less frequent random testing and fewer probation
meetings. Much can be borrowed from existing programs, but these would need to be
tailored to the needs of the California system.
References
Bazemore, G., & Maloney, D. (1994). Rehabilitating community service: Toward
restorative service in a balanced justice system. Federal Probation, 58, 24-35.
Braithwaite, J. (1989). Crime, Shame and Reintegration. Cambridge, UK: Cambridge
University Press.
Farabee, D. (2005). Rethinking rehabilitation: Why can’t we reform our criminals?
Washington, DC: AEI Press.
Grasmack, H.G., & Bryjak, G.J. (1980). The Deterrent Effect of Perceived Severity of
Punishment. Social Forces, 59, 471-491.
Harrell, A., & Roman, J. (2001). Reducing drug use and crime among offenders: The
impact of graduated sanctions. Journal of Drug Issues, 31, 207-231.
Harrell, A., & Smith, B. (1996). Evaluation of the District of Columbia Superior Court
Drug Intervention Program: Focus Group Interviews. Report to the National
Institute of Justice.
Kleiman, M. (2001). Controlling drug use and crime among drug-involved offenders:
Testing, sanctions, and treatment. In Heymann, P.H., & Brownsberger, W.N. (Eds.),
Drug Addiction and Drug Policy. Harvard University Press.
Kings County District Attorney (2007). Drug Treatment Alternative-to-Prison 16th
Annual Report. Accessed at:
http://www.brooklynda.org/dtap/DTAP%20Sixteenth%20Annual%20Report.pdf.

198

La Vigne, N.G., & Mamalian, C.A. (2004). Prisoner reentry in Georgia. Washington,
DC: Urban Institute Justice Policy Center. Accessed at:
http://www.urban.org/uploadedpdf/411170_Prisoner_Reentry_GA.pdf.
The National Center on Addiction and Substance Abuse (2003). Crossing the Bridge: An
Evaluation of the Drug Treatment Alternative-to-Prison (DTAP) Program.
Columbia University
Nichols, J. & Ross, H.L. (1990). Effectiveness of legal sanctions in dealing with drinking
drivers. Alcohol, Drugs, and Driving, 6, 33-60.
Paternoster, R. (1989). Decisions to participate in and desist from four types of common
delinquency: Deterrence and the rational choice perspective. Law and Society
Review, 23, 7-40.
Paternoster, R., Brame, R., Bachman, R., & Sherman, L.W. (1997). Do fair procedures
matter? The effect of procedural justice on spouse assault. Law and Society Review,
31, 163-204.
Reentry Policy Council (2005). Charting the Safe and Successful Return of Prisoners to
the Community. New York: Reentry Policy Council, 2005.
Rhine, E. (1993). Reclaiming Offender Accountability: Intermediate Sanctions for
Probation and Parole Violators. Laurel, MD: American Correctional Association.
Sherman, L.W. (1993). Defiance, deterrence, and irrelevance: A theory of the criminal
sanction. Journal of Research in Crime and Delinquency, 30, 445-473.
Taxman, F. (1999). Graduated sanctions: Stepping into accountable systems and
offenders. Prison Journal, 79, 182-205.
Taxman, F.S, Reedy, D., Ormond, M., & Moline, K. (2002). Break the Cycle: Year 4
Implementation. College Park: University of Maryland.
Tonry, M. (1996). Sentencing Matters. New York: Oxford University Press.
Tyler, T.R. (1990). Why People Obey the Law. New Haven: Yale University Press.

199

200

OUTCOMES AND PERFORMANCE
Chapter 11: Re-Offending and Crime Trends
Darren Urada, Ph.D. and Angela Hawken, Ph.D.

Analyses focused on re-offending (new arrests for drug, property, and violent offenses)
over a 42 month (3½-year) follow-up period in Prop 36’s first year and over a 30 month
(3½-year) follow-up period in Prop 36’s second and third years.
In one comparison, re-offending was examined in relation to the degree of offender
participation in Prop 36. Re-offending was lowest among Prop 36 offenders who
completed treatment compared to those who were referred to Prop 36 but did not enter
treatment and those who entered but did not complete treatment. New arrests for drug
offenses were substantially lower among offenders who completed treatment. Property
and violent arrests were low in all three groups.
In a second comparison, outcomes of Prop 36 as a policy were examined by comparing
re-offending among offenders in Prop 36’s first year (Prop 36-era offenders) to similar
offenders in the pre-Prop 36-era. Prop 36-era offenders had a higher rate of drug and
property arrests than the pre-Prop 36-era comparison group. Violent arrests were low in
both groups. This comparison may have been affected by differences in incapacitation
under the two policies; pre-Prop 36-era offenders were more likely to be sentenced to jail
or prison.
Patterns of re-arrests during Prop 36’s second year and third years were similar to
patterns seen in Prop 36’s first year, but drops in drug and property crime arrests were
observed between the first and second years followed by another smaller drop between
the second and third years. This trend merits continued tracking and further study to
better understand its causes.
Consistent with the comparison group differences described above, increases in drug and
property arrests were somewhat greater in California since 2001 than they were
nationally. Arrests for violent crimes fell slightly more in California than they did
nationally.
This chapter examines re-offending—new arrests for drug, property, and violent offenses—
over a 42-month follow-up period in Prop 36’s first year and over a 30-month follow-up
period in Prop 36’s second and third years.
The analyses of re-offending were twofold. First, new arrests in the follow-up period were
compared across the three groups of offenders to observe re-offending in relation to the
degree of offender participation in Prop 36. Second, Prop 36-eligible drug offenders,
including those who did and those who did not participate in treatment, were compared to a
pre-Prop 36-era group of drug offenders. This second comparison examines re-offending
under the implementation of two policy alternatives: implementation of Prop 36 policy,

201

under which drug offenders had an opportunity to accept community supervision with
treatment versus implementation of pre-Prop 36 policy, under which similar offenders were
either sentenced to prison/jail or placed under community supervision with less likelihood of
exposure to treatment.
Following both sets of analyses on the first-year cohort, these analyses were repeated on
second and third year cohorts to determine differences in re-offending between Prop 36’s
first, second, and third years.
Re-Offending in Relation to the Degree of Offender Participation in Proposition 36
The evaluation examined outcomes in the population in its first (July 1, 2001-June 30, 2002),
second (July 1, 2002-June 30, 2003), and third (July 1, 2003-June 30, 2004), years. These
populations were sorted into three mutually exclusive groups: those who were referred for an
assessment (i.e., those who accepted the opportunity to participate) but who did not receive
treatment; those who entered but did not complete treatment; and those who completed
treatment. Re-offending outcomes were adjusted for demographic, criminal history, and
drug treatment characteristics of offenders.
The purpose of this comparison was to describe re-offending in relation to the degree of
offender participation in Prop 36. Despite the effort to account for possible selection bias, it
is impossible to know precisely how the comparison serves to isolate the effect of Prop 36
itself; outcomes could be over- or under-estimated. Nevertheless the comparison is valuable
in showing the extent of re-offending among those who partially or fully complied with the
treatment requirement in Prop 36. In addition, outcomes among those who completed
treatment provide an indication of the likely maximum effect of Prop 36, at least as it was
implemented during the period of evaluation.
Prop 36 Policy Implementation versus Pre-Prop 36 Policy Implementation
This evaluation also compared the population arrested for Prop 36-eligible drug offenses in
the program’s first year and a pre-Prop 36-era population arrested for eligible offenses during
the 12-month period between July 1996 and June 19971. On most demographic and criminal
history characteristics, the Prop 36-era and pre-Prop 36-era groups were quite similar. The
Prop 36-era group, however, had a higher percentage of Hispanics, and there were some
group differences in the distribution of offenses leading to arrest (see Table 11.1). Reoffending outcomes were adjusted for background characteristics of offenders, county of
arrest, and the unemployment rate in California for the month of each offender’s arrest. The

1

Prop 36 eligibility is determined at sentencing, not at the time of arrest. UCLA used eligible convictions to
select offenders in order to obtain the best possible precision in identifying offenders eligible for Prop 36.
There are two trade-offs. First, it is possible that there were different charging practices and plea-bargaining
practices between the pre-Prop 36 and Prop 36 eras, which could potentially bias results. This bias was
mitigated to the extent possible by adjusting for differences in demographic and criminal history characteristics,
as described. Second, this focuses our analyses on following offenders with a new conviction. In particular the
subset of parolees that entered Prop 36 through a parole violation and did not have a new court conviction are
not included in these analyses. UCLA estimates that this may have excluded approximately 2% of eligible Prop
36 offenders per year. Although this is a relatively small number and they were excluded from both the Prop
36 years and comparison years, due to their parole status this group of offenders is of interest because they may
be particularly active. UCLA is working to obtain additional data on this subgroup for future research.

202

adjustment for unemployment accounts for economic conditions that might have affected reoffending.
Table 11.1 Characteristics of Prop 36-Era and Pre-Prop 36-Era Groups
1996-1997 Full
2001-2002 Full Prop
Comparison Group
36 Group
42,029
40,368

n
Sex

Male

75.4%

74.7%

Female

24.6%

25.3%

Median Age (years)

33.6

33.0

Mean Age

33.2

32.2

Asian/Pacific Islander

1.4%

2.1%

Black/African American

18.4%

16.3%

Hispanic

29.8%

32.0%

Native American

0.4%

0.5%

Other

0.5%

0.9%

Unknown

1.8%

0.4%

White

47.7%

47.7%

89.5%

91.8%

Race

Any Prior Arrests

This comparison describes re-offending period under two policy implementations: the Prop
36 policy implementation under which drug offenders had an opportunity to accept
probation/parole with treatment versus the pre-Prop 36-era policy implementation under
which those with similar offenses were either sentenced to prison/jail or placed on probation
or continued on parole with less likelihood of exposure to treatment. This comparison is
important because offenders in the Prop 36 era make a decision—whether or not to accept
Prop 36. Those who accept Prop 36 may be different from those who do not in ways that
lead to an over- or under-estimate of Prop 36 outcomes. Conversely, offenders in the preProp 36-era had no such decision to make and, thus, no opportunity to self-select. By
including all Prop 36-era offenders who met eligibility requirements at conviction and all
pre-Prop 36-era offenders who met eligibility requirements at conviction, UCLA minimized
the self-selection problem. This comparison shows how much re-offending occurred over
the 42-month period among drug offenders in the Prop 36-era and how likely re-offending
would have occurred if they had been handled under the pre-Prop 36-era policy.

203

Many offenders arrested for a Prop 36-eligible offense in the first year did not participate in
Prop 362. Some Prop 36-era non-participants (9.0%) were sentenced to jail or prison
(Longshore et al., 2007). Some of those who agreed to participate in Prop 36 (31%) did not
enter the treatment program to which they were referred (Longshore et al., 2003). On the
other hand, only some offenders in the pre-Prop 36-era (22.5%)3 were sent to jail or prison
for their eligible offense, and some (15.6%) received treatment while on probation or parole
(Longshore et al, 2004).
For these reasons, the comparison of Prop 36-era and pre-Prop 36-era eligible offenders does
not measure the effect of Prop 36 participation, nor does it show the effect of a policy under
which all offenders were sentenced to jail or prison versus an entirely different policy under
which all offenders received treatment in the community. Rather, it provides a comparison
of two time periods as two different policies were actually implemented.
Those individuals with prior or concurrent convictions that made them (or would have made
them) ineligible for Prop 36 were excluded from each offender population. Closing the preProp 36-era in June 1997 made it possible to observe re-offending over a period of 42
months during which any subsequent offending in the pre-Prop 36-era comparison group was
still subject to the pre-Prop 36-era policy.
In summary, each comparison sheds unique light on Prop 36 outcomes over an initial 42month follow-up period. The first comparison describes outcomes by Prop 36 participation
and uses treatment completers to gauge the likely maximum effect of Prop 36. The second
comparison describes outcomes of Prop 36 as a policy. These outcomes are determined by
the behavior of drug offenders who did not choose to participate in Prop 36 as well as those
who did. Effects of offender self-selection on findings thus are minimized.
Re-Offending Measure
The primary measure of re-offending was based on new arrests that occurred during the
period after the Prop 36-eligible conviction. Arrests are an imprecise measure of offending
because many offenses are undetected by law enforcement and because an officer’s arrest
decision, given detection of a possible offense, is, in many cases, discretionary (Blumstein,
2002). Moreover, occurrence of an arrest does not necessarily mean that the person
committed a crime. On the other hand, the offense for which an arrestee is later charged or
convicted depends on a series of additional discretionary decisions by prosecutors and judges
(Blumstein & Cohen, 1979; Forst, 2002), and the disposition of an arrest (e.g., charge
dismissed, defendant acquitted, or defendant convicted) is often missing from criminal
justice records. New arrests, therefore, are the most appropriate indicator of re-offending for
the purpose of group comparison. Arrests come “closer to the crime” than other data
available in criminal justice records and are most commonly used by criminologists to
measure re-offending (Maltz, 2001).
2

UCLA examined records for drug offenders who were arrested for Prop 36-eligible offenses but did not
participate in Prop 36. Of offenders with dispositions, some (7%) were acquitted or had their cases dismissed.
Some entered drug court (6%) or were routed to a “deferred entry of judgment” program (4%). Most of those
with a conviction were sentenced to a jail term (56%), usually followed by probation.
3
According to DOJ records, 9.3% were sent to jail for felony drug offenses and 6.4% for misdemeanor drug
offenses; 6.7% were sent to prison for felony drug offenses and 0.1% for misdemeanor drug offenses.

204

Separate measures were used to examine the percentage of offenders with a new arrest for a
drug offense, property offense, and violent offense. For each offense type, felonies and
misdemeanors were examined separately and in combination. The time period in which reoffending could occur was 42 months after the Prop 36-eligible conviction. Violations of
probation or parole were not counted unless the violation was a new offense resulting in
arrest. Issuance and execution of warrants were not counted. Accordingly, measures of reoffending reflected new criminal activity. The analysis covered property and violent arrests
as well as drug arrests because drug-related crime could have carry-over effects on incomegenerating property crime or violence associated with drug markets.
Re-Offending among Prop 36 Participants
New arrests were least common among Prop 36-era offenders who completed treatment. As
shown in Figure 11.1, the 42-month drug arrest rate was 61.5% among referred offenders
who did not receive treatment, 65.1% among offenders who entered but did not complete
treatment, and 46.9% among those who completed treatment. Property arrests were similar
for offenders who did not receive treatment (19.6%) and those who entered but did not
complete treatment (18.5%), but lower for those who completed treatment (11.8%). As with
drug and property arrests, violent arrests were least common among treatment completers,
but such arrests were uncommon in all groups and differences therefore were small.

Figure 11.1
New Arrests During 42 Months After Offense
Prop 36 Offenders, July 2001 – June 2002
(N =17,519)

Percent of offenders

100

80
61.5

65.1

60
46.9

40
19.6

20

18.5
11.8
6.4

5.9

4.1

0
New drug arrest
Referred but untreated
(N = 6,954)

New property arrest
Entered but did not complete treatment
(N = 7,611)

205

New violent arrest
Completed treatment
(N = 2,954)

When new arrests were separated into felonies and misdemeanors, these patterns recurred.
See Figures 11.2 and 11.3.

Figure 11.2
New Felony Arrests During 42 Months After Offense
Prop 36 Offenders, July 2001 – June 2002
(N = 17,519)
Percent of offenders

100

80

60

44.6 47.0
40

32.3
16.3 14.8

20

9.3
3.7

2.9

2.3

0
New drug arrest
Referred but untreated
(N = 6,954)

New property arrest
Entered but did not complete treatment
(N = 7,611)

New violent arrest
Completed treatment
(N = 2,954)

Figure 11.3
New Misdemeanor Arrests During 42 Months After Offense
Prop 36 Offenders, July 2001 – June 2002
(N = 17,519)
Percent of offenders

100

80

60

40

32.1 33.6
23.4

20

5.4

5.4

3.4

2.6

2.8

1.0

0
New drug arrest
Referred but untreated
(N = 6,954)

New property arrest
Entered but did not complete treatment
(N = 7,611)

New violent arrest
Completed treatment
(N = 2,954)

Re-Offending Under Prop 36-Era and Pre-Prop 36-Era Policies
The percentage of offenders with a new drug arrest was higher in the Prop 36-era than in the
pre-Prop 36-era. As shown in Figure 11.4, 55.2% of offenders in the Prop 36-era and 48.9%
in the pre-Prop 36-era had a new drug arrest during the 42-month follow-up period. Arrests
206

for property crimes were also somewhat higher in the Prop 36-era group. Arrests for violent
crimes were similar and low in both groups.

Figure 11.4
New Arrests During 42 Months After Offense
Prop 36 Year One vs Pre- Prop 36 Comparison Group

Percent of offenders

100

80

55.2

60
48.9

40
20.1

20

14.1
5.6

6.4

0
New drug arrest

New property arrest

Comparison offenders
(N = 42,029)

New violent arrest

Prop. 36 eligible offenders
(N = 40,368)

When arrests were separated into felonies and misdemeanors, the patterns were generally the
same. However, felony property arrests were more common in the Prop 36-era group than in
the pre-Prop 36-era comparison group (see Figures 11.5 & 11.6).

Figure 11.5
New Felony Arrests During 42 Months After Offense
Prop 36 Year One vs Pre-Prop 36 Comparison Group

Percent of offenders

100

80

60

40

39.5
34.2

16.7

20
10.6

3.1

3.8

0
New drug arrest

New property arrest

Comparison offenders
(N = 42,029)

New violent arrest

Prop. 36 eligible offenders
(N = 40,368)

207

Figure 11.6
New Misdemeanor Arrests During 42 Months After Offense
Prop. 36 Year One vs Pre-Prop. 36 Comparison Group

Percent of offenders

100

80

60

40
23.9

28.0

20
5.0

5.0

2.6

2.4

0
New drug arrest

New property arrest

Comparison offenders
(N = 42,029)

New violent arrest

Prop. 36 eligible offenders
(N = 40,368)

Pre-Prop 36-era drug offenders were more likely than Prop 36-era drug offenders to be
sentenced to jail or prison following arrest for the eligible offense. Accordingly, pre-Prop
36-era offenders had less opportunity to re-offend because, during the follow-up period, they
were more likely to be in custody for part or all of the period. This difference in sentencing
is one aspect of the policies being compared. Hence, for a clear look at outcomes of these
policies, there should be no adjustment for it.
First, Second, and Third Year Re-arrests, 30-Month Follow-up
Re-arrest trends were examined for offenders who were referred to Prop 36 in Years 1, 2,
and 3. Patterns of re-arrests in Years 2 and 3 were very similar to those in Year 1.
Treatment completers had far fewer re-arrests than offenders who were referred but not
treated, and those who started but did not complete treatment (see Figure 11.74).

Prop 36’s first three years were also compared to the pre-Prop 36-era group over a 30-month
follow-up period. Patterns of re-arrests during Prop 36’s second year and third years were
similar to patterns seen in Prop 36’s first year, but small drops in drug and property crime
arrests were observed between the first and second years followed by another (smaller) drop
between the second and third years. This trend clearly merits continued tracking and further
study to better understand its causes. Across crime categories, Prop 36-era offenders were
somewhat more likely to be re-arrested compared to pre-Prop 36-era offenders (see Figure
11.85).
4

Year one numbers in Figure 11.7 are very similar but not identical to year one statistics found in Figure 4.1 of
the report UCLA released in 2007 due to updated data and methods. All differences are less than 1.5%.
5
Year one numbers in Figure 11.8 are very similar but not identical to year one statistics found in Figure 4.4 of
the report UCLA released in 2007 due to updated data and methods, and different comparison group years.
1996-1997 was used as the comparison in this report while 1997-1998 was used in the previous report. As

208

Figure 11.7
New Arrests During 30 Months After Offense
Prop 36 Offenders, Years 1-3

80

35.2

4.7
4.0
2.9

8.3

4.9
4.7
3.4

8.8

16.4
15.6

16.7
15.6
5.1
4.7
3.0

16.4
15.4
9.2

40
20

54.6
57.2

55.2
58.3
38.7

41.6

60

56.9
60.1

Percent of offenders

100

0
New drug
New
arrest
property
arrest

New
violent
arrest

New drug
New
arrest
property
arrest

Year One
Referred but untreated

New
violent
arrest

New drug
New
arrest
property
arrest

Year Two

New
violent
arrest

Year Three

Entered but did not complete treatment

Completed treatment

Figure 11.8
New Arrests During 30 Months After Offense
Prop 36 Years 1-3 vs Pre-Prop 36 Comparison Group

Percent of offenders

100

80

60
42.7

49.0 46.6
46.1

40

20

11.5

16.4 16.1 16.0
4.4

0

New drug arrest

New property arrest

5.0

5.1

4.9

New violent arrest

Comparison offenders

Prop. 36 year one eligible

Prop. 36 year two eligible

Prop. 36 year three eligible

discussed earlier in the chapter, the change in comparison group years was necessary to provide an accurate
comparison for the 42-month follow-up analyses. For consistency, this comparison group was also used in the
30-month analyses presented here. Year one numbers differ by less than 1% in all cases.

209

Statewide Crime Trends
Violent and property crime trends for California and the nation are shown in Figure 11.9 and
Figure 11.10. The violent crime arrest rate per 100,000 fell 28% in California between 1994
and 2005, the property arrest rate fell by 37%, and the drug arrest rate fell by 1%.
Nationwide, between 1994 and 2005 the violent crime rate per 100,000 fell 30%, the
property crime rate fell by 26%, and the drug arrest rate increased by 20%. Between 2001
(the year that Prop. 36 was implemented) and 2005, violent crime in California fell 12%
(nationwide, violent crime fell 9%), there has been a 6% increase in property index crimes
(nationwide, property crime remained stable over the same period), and drug arrests have
increased by 21% (nationwide, drug arrests increased by 14%).
Figure 11.9
California and U.S. Violent Index Crime Rate (1994 – 2005)
700

Crime rate

600
500
400
300
200
100
0
1994

1995

1996

1997

1998

CA Violent Index (rate per 100,000)

1999

2000

2001

2002

2003

2004

2005

U.S. Violent Crime Index (rate per 100,000)

Note: Data are from the FBI Uniform Crime Reports as prepared by the National Archive of Criminal Justice
Data. The shaded area indicates the Prop 36-era. Prop 36 was implemented on the fiscal year (July 2001),
whereas crime data is reported on the calendar year.

Here we provide a more detailed analysis of statewide drug arrests in California from the
California Department of Justice. Many factors influence statewide drug arrests. The
analysis below is descriptive only and does not isolate the causal effect of Prop 36 on
statewide drug arrests.
Figure 11.12 shows statewide felony and misdemeanor drug arrests in California from 1997
to 2005. There has been an increase in felony drug arrests statewide. The late 1990s saw a
steady decline in felony drug arrests; the implementation of Prop 36 coincided with a
reversal of this pattern. Total felony drug arrests have increased by 30% since Prop 36 was
implemented. Misdemeanor arrests have increased by 23% since Prop 36 was implemented.
The increase in drug arrests was primarily due to an increase in arrests for methamphetamine
use.

210

Figure 11.10
California and U.S. Property Index Crime Rate (1994 – 2005)
900

Crime rate

800
700
600
500
400
300
200
100
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
CA Property Index (rate per 100,000)

U.S. Property Crime Index (rate per 100,000)

Note: Data are from the FBI Uniform Crime Reports as prepared by the National Archive of Criminal Justice
Data. The shaded area indicates the Prop 36-era. Prop 36 was implemented on the fiscal year (July 2001),
whereas crime data is reported on the calendar year.

Figure 11.11
California and U.S. Drug Crime Trends (1990 – 2005)
1,200

Crime rate

1,000
800
600
400
200
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
California Drug Crime (rate per 100,000)

U.S. Drug Crime (rate per 100,000)

Note: Data are from the FBI Uniform Crime Reports as prepared by the National Archive of Criminal
Justice Data. The shaded area indicates the Prop 36-era. Prop 36 was implemented on the fiscal year (July
2001), whereas crime data is reported on the calendar year.

211

Figure 11.12
California Misdemeanor and Felony Drug Arrests

Number of arrests

180,000
160,000
140,000
120,000
100,000
80,000
60,000
40,000
20,000
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Felony arrests

Misdemeanor arrests

Data source: California Department of Justice

Conclusion
Findings in this chapter were based on two types of comparisons. The first described
outcomes among Prop 36’s first-year participants in relation to the degree of offender
participation in Prop 36. The second comparison described outcomes of Prop 36 as a policy
among drug offenders who did and did not choose to participate in Prop 36. Both
comparisons focused on Prop 36 outcomes over a 42-month follow-up period for the first
year. In addition, the results were replicated over 30-month follow-up periods for both the
first and second year in order to assess changes between the first two years.
Outcomes among offenders who completed treatment provided an indication of the likely
maximum short-term effect of Prop 36 in the first year. The analysis found that less than
half of treatment completers had a new drug arrest during the 42-month follow-up period,
whereas nearly two-thirds of those who did not complete treatment were re-arrested.
In the comparison of the two policy alternatives, arrests were higher among Prop 36-era
offenders than in a similar group of pre-Prop 36-era offenders on drug offenses and property
offenses. Re-offending was low and similar across groups for felony and misdemeanor
violent arrests. By including all Prop 36-era offenders arrested for an eligible drug crime and
all pre-Prop 36-era offenders arrested for a drug crime that would have been eligible, this
comparison showed how much re-offending occurred over a 42-month follow-up period
among drug offenders in the Prop 36-era and how much likely would have occurred if they
had been handled under the pre-Prop 36-era policy.
There was a clear pattern in the findings, with the lowest re-offending outcomes evident
among those who completed treatment. This finding is typical of studies comparing such
groups (e.g., Inciardi et al., 2004; Prendergast et al., 2003).

212

Findings were affected by differences in incarceration under Prop 36-era and pre-Prop 36-era
policies. Offenders who are incarcerated are unable to be re-arrested for new crimes.
Outcomes were very similar across Prop 36’s first three years, but there was a trend toward
fewer offenders being arrested in each successive cohort in the first three years of Prop 36
implementation. This potentially positive trend will be monitored closely and studied further
by UCLA to see whether it continues and whether maturation of implementation practices
may have played a role.
Prop 36 cannot be causally linked to statewide arrest trends using the data available.
However, statewide arrest trends are generally consistent with the pattern that would be
expected based on the comparison group differences described above overlaid upon national
trends. In the statewide arrest trends, increases in arrests since 2001 were somewhat higher
than national increases in drug and property offenses (but notably not violent ones6), the
same crime categories where the largest increases were seen in the comparison group
analyses. Because new offenders become eligible for Prop 36 each year and add to the
ranks, the continuing increase in statewide arrests is not inconsistent with the finding that
fewer Prop 36 eligible offenders were arrested in each successive cohort. Should this trend
continue, statewide arrest trends should level off. However, the increase in drug arrest rates
since 2001 also occurs at the same time as a national increase, suggesting that larger issues
are contributing to the trend.
It is important to note that outcomes described above are a reflection of Prop 36 policy as
written and of Prop 36 treatment and supervision as delivered. Under Prop 36 policy,
eligible drug offenders may or may not choose to participate in Prop 36. Among those who
did choose Prop 36, the degree of participation, as indicated by treatment entry and
completion, varied widely between offenders. Outcomes might have been different if policy
and implementation practices were different.

References
Blumstein, A. (2002). Prisons: A policy challenge. In J. Q. Wilson, & J. Petersilia
(Eds.), Crime: Public policies for crime control (pp. 451-482). Oakland, CA:
Institute for Contemporary Studies Press.
Blumstein, A., & Cohen, J. (1979). Estimation of individual crime rates from arrest
records. Journal of Criminal Law and Criminology, 70, 561-585.
Forst, B. (2002). Prosecution. In Wilson, J.Q. & Petersilia, J. (Eds.), Crime: Public
policies for crime control. Oakland, CA: Institute for Contemporary Studies Press.
Inciardi, J.A., Martin, S.S., & Butzin, C.A. (2004). Five-year outcomes of therapeutic
community treatment of drug-involved offenders after release from prison. Crime &
Delinquency, 50, 88-107.
6

One untested hypothesis is that as Prop 36 has removed drug offenders from overcrowded jails and prisons,
space has been created to incarcerate more offenders sentenced for more-serious than would be possible in the
absence of Prop 36, creating a larger drop in violent crime than would have been possible in the absence of
Prop 36.

213

Longshore, D., Urada, D., Evans, E., Urada, D., Teruya, C., Hardy, M. Hser, Y.-I.,
Prendergast, & Ettner, S. (2003). Evaluation of the Substance Abuse and Crime
Prevention Act: 2002 report. Sacramento, CA: Department of Alcohol and Drug
Programs, California Health and Human Services Agency
Longshore, D., Urada, D., Evans, E., Hser, Y.-I., Prendergast, M., Hawken, A., Bunch,
T., & Ettner, S. (2004). Evaluation of the Substance Abuse and Crime Prevention
Act: 2003 report. Sacramento, CA: Department of Alcohol and Drug Programs,
California Health and Human Services Agency
Longshore, D., Urada, D., Hawken, A., Anglin, M.D., Conner, B.T., Evans, E., Hser, Y.I., Prendergast, M., Hiromoto S., Du D., & Bunch T. (2005). Evaluation of the
Substance Abuse and Crime Prevention Act: 2005 report. Sacramento, CA:
Department of Alcohol and Drug Programs, California Health and Human Services
Agency
Maltz, M.D. (2001). Recidivism. Orlando, Fl: Academic Press, Inc. Retrieved on
February 25, 2005, from www.uic.edu/depts/lib/forr/pdf/crimjust/recidivism.pdf.
Prendergast, M., Hall, E., & Wexler, H. (2003). Multiple measures of outcomes in
assessing a prison-based drug treatment program. Journal of Offender
Rehabilitation, 37, 65-94.

214

Chapter 12: Proposition 36 Benefit-Cost Analysis
Angela Hawken, Ph.D., Douglas Longshore, Ph.D., Darren Urada, Ph.D., Jia Fan, M.S.,
and M. Douglas Anglin, Ph.D.

UCLA conducted three studies assessing the cost implications and benefit-cost ratios of Prop
36. Each showed that Prop 36 yielded cost savings to state and local governments.
Study 1 extended the baseline and follow up periods used in UCLA’s earlier cost report from
30 months to 42 month. Here, costs for a pre-Prop 36-era comparison group and for all firstyear Prop 36-eligible offenders found a net savings of $1,977 per offender (N = 61,609) over
a 42 month period, yielding a benefit-cost ratio of nearly 2 to 1. In other words, $2 was saved
for every $1 invested.
Study 2 used first year Prop 36 participants who were referred to the program. Prop 36
participants who completed treatment achieved a benefit-cost ratio of approximately 4 to 1
over a 42 month period, indicating that “completers” saved $4 for every $1 allocated.
Study 3 examined follow-up costs for succeeding year as the policy matured. Over a 30
month follow up period, the costs for jail, probation, parole, and treatment have remained
stable from year to year. Prison costs and costs for arrest and convictions have steadily
declined over the first 3 years.
Two conclusions follow from the cost analyses: Prop 36 substantially reduced incarceration
costs and resulted in greater cost savings for some eligible offenders than for others.
The purpose of the Prop 36 benefit-cost analysis was to examine: 1) overall costs to state and
local government for drug offenders eligible for Prop 36, 2) cost patterns based on the degree
of Prop 36 treatment participation by offenders and 3) possible changes in cost outcomes for
consecutive Prop 36 cohorts (Prop 36 eligible in year 1, year 2, and year 3 after the
implementation of the law).
Study 1 calculated the benefit-cost ratio attributable to Prop 36 as a policy, that is, as a
change in law that applied to all offenders throughout the state, regardless of the degree of
offender participation. Study 2 examined variation in benefit-cost ratios in relation to
offenders’ degree of participation in Prop 36 treatment. This study assessed benefit-cost
outcomes for offenders who accepted the drug treatment option at conviction (i.e., accepted
referral to the Prop 36 program), whether or not they entered drug treatment, and whether
they completed the planned treatment. A particular focus of Study 2 was the benefit-cost
ratio for those who completed drug treatment (completers). Study 1 and Study 2 were based
on Prop 36’s first-year population of eligible offenders and covered a 42-month baseline
period and a 42-month follow-up period from the eligible conviction.
Study 3 examined the potential change in benefit-cost ratio estimates from the first to the
third year of Prop 36 to determine whether cost outcomes changed as Prop 36 matured.
Study 3 used 30-month baseline and follow-up periods. All three studies used the “taxpayer
215

perspective,” focusing on costs to state and local governments. Results are expressed in
average cost or savings per offender. Furthermore, all costs were adjusted to 2004 dollars to
allow standardization across multiple years and comparison with previous reports.
The three studies reported here differ from the original benefit-cost analysis of Prop 36
included UCLA’s 2007 report SACPA Evaluation: Final Report in five ways:
1. The follow-up periods for studies 1 and 2 are 12 months longer than those in the
original cost report.
2. To permit a longer follow-up period, a new pre-Prop 36 comparison group was
constructed. The new comparison group consists of all individuals convicted
between July 1, 1996 and June 30, 1997 who would have been eligible for Prop 36
had the law been in place.
3. Due to data lags for the third year cohort, a 30 month follow-up period was used for
study 3.
4. As data were not provided by EDD, the earnings module is excluded from the three
studies in this chapter. The original benefit-cost analysis included an earnings
module which resulted in a slight increase in savings attributable to Prop 36 as a
result of increased tax revenues collected on earnings.
5. Health outcomes are omitted for study 3 due to problems with data matching.
This report provides the essential findings and the subsequent conclusions and
recommendations from the three studies. The savings and costs reported across the seven
areas (modules) represent the net savings (or costs) that can be attributed to Prop 36. This
report also summarizes the analytic process undertaken to provide valid and consistent data,
appropriate analysis, and suitable adjustments for the cost components under consideration.
Background
Prop 36 was enacted by California voters as a statewide policy that changed the course of
criminal justice processing for all eligible offenders, whether or not they chose to participate
in the program. The policy also affected all service entities that interact with the pool of
eligible offenders. The most rigorous and conservative scientific approach required the
construction of a comparison group. Since the most-preferred study design, with offenders
randomly assigned to either Prop 36 or non- Prop 36 interventions, was not possible, a
comparison group was constructed by selecting similar offenders convicted of Prop 36eligible crimes from a period before Prop 36 was implemented, referred to as the pre- Prop
36-era. UCLA compared the total statewide costs for drug offenders eligible for Prop 36
during its first year (Prop 36-era N = 61,609) to total statewide costs for a selected
comparison group of drug offenders before Prop 36 was initiated (pre-Prop 36-era N =
68,543)1. The analytic approach used is a significant improvement to that of cost studies
limited to single-group, pre/post-designs, such as the California Drug and Alcohol Treatment
Assessment (CALDATA; Gerstein et al., 2005) and the California Treatment Outcome
1

While the pre-Prop 36-era and Prop 36-era groups had different sample sizes, the samples were used only to
obtain per-offender costs in the seven areas. Once these costs were determined, the calculation of total costs
was rebased to the Prop 36 sample size.

216

Project (CalTOP; Hser et al., 2005). The Prop 36 benefit-cost analysis also improved on
such studies by using official records for data sources, thus removing the need to rely
primarily on subject self-report. Finally, the study used lengthy baseline and follow-up
periods, thus limiting the effects of “regression to the mean,” which can spuriously inflate
post-intervention benefits2.
The benefit-cost analysis comprises three studies. Study 1 calculated the net savings (or
costs) and benefit-cost ratio attributable to Prop 36 as a policy applied statewide to all
eligible offenders. Study 2 examined variation in cost ratios in relation to offenders’ degree
of participation in Prop 36. A particular focus of Study 2 was the benefit-cost ratio for
offenders who completed their Prop 36 drug treatment program. Study 1 and Study 2 were
based on Prop 36’s first-year population and covered a 42-month baseline period and a 42month follow-up period from the eligible conviction date. Study 3 examined the change in
net savings (or costs) for the first, second, and third years of Prop 36 to assess if maturity of
the policy may have changed cost outcomes. This analysis made it possible to compare more
precisely each year’s costs to the $120 million annual Prop 36 allocation provided for drug
treatment and other services3. Study 3 replicates the first-year analysis, and confers greater
confidence in the results of Study 1. However, since the second-year and third-year cohorts
were drawn from a more recent period than the first-year cohort, there was a shorter period
available for follow-up. As a result, Study 3 used 30-month baseline and follow-up periods
around the Prop 36-eligible conviction in order to capture equal periods for comparison of
the first-, second-, and third-year Prop 36 offenders. As noted earlier, all three studies used
the taxpayer perspective, in which the focus is on costs to state and local governments. All
costs were adjusted to 2004 dollars to permit comparisons with previous reports, using the
consumer price index or, where appropriate, the medical price index4. Costs have been
rounded to the nearest dollar.
The findings, conclusions and recommendations, and analytic methods are summarized in
this report and its appendices. Study findings are presented in the first section, followed by
conclusions and recommendations. A final section describes the analytic design employed,
the data used, and methodological techniques applied.
Prop 36 First-Year Cost Analysis (Study 1)
Study 1 compared offenders eligible for Prop 36 with a pre-Prop 36-era group of offenders
who would have been eligible for Prop 36 under the law’s provisions5. The purpose of this
2

“Regression to the mean” refers to the tendency of individuals with below-average problems and costs in one
period to have more problems and higher costs in the next period, and vice versa for those with above-average
problems and costs. Many individuals enter treatment when they have the most problems (Harwood et al.,
2002). This is especially true for individuals entering treatment under a court mandate following a conviction.
It is possible that, in the absence of the treatment intervention, the client would have improved on a number of
outcome measures, in other words, part of the beneficial pre/post change would have been observed anyway.
3
The study conservatively assumes programmatic costs to be $117 million under Prop 36, excluding $3 million
of the $120 million annual allocation that was used to cover state-level administrative expenses.
4
U.S. Department of Labor, Bureau of Labor Statistics. Consumer Price Index: 1913–2005, [data file].
Washington, DC: www.bls.gov/data.
5
Offenders were drawn from official California Department of Justice records on arrests and convictions with
subsequent computerized eligibility screening. These numbers are larger than those estimated in prior reports,
which were obtained from stakeholder surveys or the centralized SRIS derived from county inputs.

217

analysis was to calculate the cost attributable to Prop 36 as a policy. The Prop 36.-era group
was the population of adults (18 years or older) who were, during Prop 36’s first year (July
1, 2001 to June 30, 2002), convicted of a Prop 36-eligible offense with no concurrent nondrug offense or other circumstance that made them ineligible. Study 1 uses a 42-month
follow-up and baseline (follow-back) period. The 42-month follow-up period for each Prop
36-era offender ended on or before December 31, 2005. The comparison, or pre-Prop 36era, group6 was drawn from a population of adults convicted of an offense for which they
would have been Prop 36-eligible had they been convicted after Prop 36 was implemented,
with no concurrent non-drug offense or other circumstance that would have made them
ineligible. This population of offenders was convicted between July 1, 1996 and June 30,
1997. The 42-month follow-up period for all comparison offenders ended on or before
December 31, 2000, at least six months before Prop 36 may have begun to affect the
involved systems. Findings covered the 42- month baseline and follow-up periods beginning
with the date of each offender’s conviction.
This section first reports the difference-in-differences (DID), calculated as the difference
between (1) the Prop 36-era group’s pre-conviction and post-conviction difference in costs
and (2) the pre-Prop 36-era group’s pre-conviction and post-conviction difference in costs
(see appendix). This yields a DID average-cost per offender in each cost area. Outcomes in
each module are documented and a summary of the cost profile of Prop 36-related costs or
savings across all seven areas is provided.
Cost per Offender
The estimates below reflect regression-adjusted average (mean) savings or costs per offender
for the pre-Prop 36-era and Prop 36-era groups in each cost category. Costs were calculated
based on events, as captured in state administrative databases, multiplied by the costs
associated with the event, as determined from data or published sources.

The figures report costs in the baseline period; costs in the follow-up period; differences
from baseline to follow-up for each group; and the DID between groups (costs are positive
numbers and savings are negative numbers). The full assumptions and statistical techniques
underlying these estimates are provided in the Research Methods section.
Prison
Prison costs are shown in Figure 12.1. Cost per offender increased by $4,312 over a 42month baseline period for the Prop 36-era group and by $8,614 for the pre-Prop 36-era
group, which led to a DID prison-cost savings of $4,302 during Prop 36. This means that
prison costs in California were $4,302 lower per offender for the 42-month follow-up period
than what would have been had Prop 36 not been implemented. For the 61,609 offenders
eligible for Prop 36 in its first year, the total savings to the state in prison costs over a 42
month period were $265 million.

6

Because the pre-Prop 36-era comparison group was, of necessity, drawn from a different period, it is
technically known as a time-lagged comparison group.

218

Figure 12.1
Prison Costs
$14,000
$12,082
$12,000

Dollars per offender

$10,000
$8,775

$8,614
$8,000

$6,000
$4,463
$4,000

$4,312

$3,467

$2,000

$0
Comparison

Prop 36
Pre

Post

Difference

DID = -$4,302

Notes: Data for number of days served in prison are from the Offender-Based Information System.
Cost of a prison day ($84.74) was obtained from the California Department of Corrections (2005).
Because the number of prison days avoided by Prop 36 offenders exceeded a full census of a midsize facility, the average cost of a prison day was used rather than the marginal cost.

Jail
Jail costs are shown in Figure 12.2. Cost per offender increased by $2,106 over baseline
during the Prop 36-era and by $3,968 for the pre-Prop 36-era group, a DID jail cost savings
of $1,862. This means that jail costs under Prop 36 were $1,862 lower per offender during
the 42-month follow-up period than would have been expected in the absence of Prop 36.
Total savings in jail costs to counties for first-year Prop 36 offenders were $115 million.
Probation
The cost of probation supervision is shown in Figure 12.3. Cost per offender increased by
$1,798 over baseline for the Prop 36-era group and by $1,072 for the pre-Prop 36-era group,
which led to a DID probation supervision cost increase of $727. This result means that
probation costs during the Prop 36-era were $727 higher per offender for the 42-month
follow-up period than would have been expected in the absence of Prop 36. Total additional
cost to the counties for probation was $45 million.

219

Figure 12.2
Jail Costs
$7,000
$6,207
$6,000

Dollars per offender

$5,000
$4,212

$3,968
$4,000

$3,000
$2,239

$2,106

$2,106

$2,000

$1,000

$0
Comparison

Prop 36
Pre

Post

Difference

DID = -$1,862

Notes: Data for number of days sentenced to jail are from the California Department of Justice
Automated Criminal History System. The 2005 ADP County Survey was used to adjust to actual days
served. Cost of a jail day by county was obtained from the County Survey and the 2003 California
Board of Corrections Survey.

Figure 12.3
Probation Costs
$3,000

$2,778

$2,500

Dollars per offender

$2,171
$2,000

$1,798

$1,500
$1,099

$1,072

$980

$1,000

$500

$0
Comparison

Prop 36

DID =$727
Pre

Post

Difference

Notes: Data for number of days on probation is from sentencing records in the California Department of
Justice Automated Criminal History System. Cost of a probation day was obtained by county from the
2005 ADP County Survey.

220

Parole
The cost of parole supervision is shown in Figure 12.4. Cost per offender increased by $277
over baseline during the Prop 36-era and by $573 for the pre-Prop 36-era group, a DID
parole supervision cost decrease of $296. This means that parole costs under Prop 36 were
$296 lower per offender for the 42-month period than what would have been expected in the
absence of Prop 36. This difference was expected, given the lower number of prison days
(see Figure 7.1) served by Prop 36 offenders. Parole cost savings to the state under Prop 36
were $18 million.

Figure 12.4
Parole Costs
$1,200
$964

Dollars per offender

$1,000

$800

$745
$573

$600

$468
$400

$391
$277

$200

$0
Comparison

Prop 36

DID =-$296
Pre

Post

Difference

Notes: Data for number of days on parole is from sentencing records in California Department of Justice
Automated Criminal History System. Cost of a parole-day ($9.21) is from the California Department of
Corrections (2004).

Arrests and Convictions
Arrest and conviction costs are shown in Figure 12.5. Although both costs declined for both
groups in the 42-month follow-up period, they did not decrease by as much for the Prop 36era group. This was due in part to the longer time that offenders in the pre-Prop 36-era
group were “off the street” during the follow-up period due to incarceration. Since offenders
who are incarcerated are unavailable to be re-arrested in the community, these differences in
street time would be expected to reduce re-arrests and convictions to a greater degree in the
pre-Prop 36-era group than in the Prop 36-era group7. Costs per offender decreased by $443
7

Every judicial decision to place an offender on probation contains a degree of risk of re-offending in the
community. This is also true when inmates are paroled. In general, any population of offenders under legal
supervision has rates of re-offending that increase in proportion to time on the street. Many policy studies on
the benefit-cost ratio of incapacitation (incarceration) have assessed the “balance point” between the high cost
of incarceration and the greater risk of re-offending under lower-cost community supervision.

221

relative to baseline levels for the Prop 36-era group and by $2,418 for the pre-Prop 36-era
group. DID arrest-and-conviction costs were $1,975 higher for the 30-month follow-up
period than what would have been anticipated had Prop 36 not been implemented, resulting
in a total increase of $122 million in criminal justice processing costs.
Figure 12.5
Arrest and Conviction Costs
$10,000
$8,272

Dollars per offender

$8,000
$6,000

$7,050

$6,607

$5,854

$4,000
$2,000
$0
Comparison

Prop 36

-$443

-$2,000
-$2,418
-$4,000
Pre

Post

Difference

DID = $1,975

Notes: Numbers of arrests and convictions are from sentencing records in the California Department of Justice
Automated Criminal History System. Costs for crime were adjusted from Miller and colleagues (1996) and
French (2005).

Drug Treatment
Drug-treatment costs are shown in Figure 12.6. Cost per offender increased by $1,545
over baseline for the Prop 36-era group and by $429 for the pre-Prop 36-era group, a DID
increase of $1,116 per offender, resulting in $69 million more in treatment costs than
what would have been anticipated had Prop 36 not been implemented.
Healthcare
Healthcare costs are shown in Figure 12.7. Costs per offender increased by $1,289 for the
Prop 36-era group and by $622 for the pre-Prop 36-era group. Such costs were $667 higher
per offender for the 42-month follow-up period than would have been anticipated had Prop
36 not been implemented. Healthcare costs to the state increased by $41 million under Prop
36.

222

Figure 12.6
Drug-Treatment Costs
$2,500
$2,217

Dollars per offender

$2,000
$1,545
$1,500
$993
$1,000
$672

$564
$429

$500

$0
Comparison

Prop 36
Pre

Post

Difference

DID = $1,116

Notes: Data for treatment days by modality are from CADDS. Per-diem costs are from Ettner and
colleagues (2006) adjusted to 2004 dollars.

Figure 12.7
Healthcare Costs
$3,500
$3,052

Dollars per offender

$3,000
$2,500
$1,981

$2,000
$1,500

$1,763
$1,359

$1,289

$1,000
$622
$500
$0
Com parison

Prop 36
Pre

Post

Difference

Notes: Data for healthcare costs are from DHS Medi-Cal/Medicaid files.

223

DID =$667

Prop 36 Overall Cost-Offsets
Figure 12.9 shows a summary of Prop 36 DID costs over all areas examined. The zero line
is interpreted as cost neutral. Bars above the line represent cost increases and bars below the
line represents cost savings. There was a total DID cost savings of $1,977 per offender
under Prop 36 over the 30-month follow-up period8.

Figure 12.9
DID Cost Summary for Study 1
$3,000
$1,974
$2,000
$1,116
$727

Dollars per offender

$1,000
Prison

Jail

$667
Parole

Total

$0
Probation
-$1,000
-$2,000

-$296

-$1,862

Arrest and
Conviction

Treatment

Health

-$1,977

-$3,000
-$4,000
-$4,303
-$5,000

Study 1 allowed the calculation of a total DID cost for the population of 61,609 offenders in
Prop 36’s first year. Before turning to the calculation of the benefit-cost ratio, it must be
noted that the initial year required a massive ramp-up effort by the involved county systems.
The expansion of existing provider contracts and the development and awarding of new
contracts was, in many cases, a lengthy process. In addition, during this year, state and
county governments were coping with the overall budget constraints of a faltering economy.
In some counties, non-recurring funds were used in ways that allowed savings to accrue to
the allocated Prop 36 funds. These savings could then be carried forward into future years.
In the first year of the study, 55 of the 58 counties reported9 a total expenditure of Prop 36
funds of about $85 million, an amount less than actually spent. Using this figure would have
produced a spuriously high benefit-cost ratio for the first year. Accordingly, UCLA used an
estimate of Prop 36 operation costs ($120 million less $3 million used for state
administrative costs) as a conservative estimate of expenditures (a figure that stabilized in the
subsequent years of Prop 36).
To determine the benefit-cost ratio per offender for the first year, total costs over the 42month period (expressed as a negative number, which represents savings) are multiplied by
the total number of offenders convicted of a Prop 36-eligible offense during the first year of
Prop 36 (N = 61,609)10. From this total, the $117 million actually allocated for
8

Most of these savings accrued in the first 12 months of this period, although savings continued to accrue over
the remaining 18 months of the 30-month period. See results of Study 3.
9
Figures cited are from SRIS.
10
Earlier UCLA reports estimated the number of eligible offenders from the Stakeholder Survey for the first
year and from SRIS for the second (reported by county lead agencies). The cost analysis improved on these
estimates by using official DOJ records.

224

programmatic costs is subtracted to avoid “double counting” costs that had already been paid
for via Prop 36 expenditures ($120 million less the $3 million used in Prop 36
administration). The resulting sum is divided by the $120 million allocated for first-year
Prop 36 costs. In brief, the benefit-cost ratio reported is the total savings net of
programmatic costs derived from Prop 36, divided by the $120 million allocation11.
For Study 1, UCLA estimated a benefit-cost ratio of 1.99:1, meaning that nearly $2 was
saved under Prop 36 for every $1 allocated to fund the program.
Prop 36 Drug Treatment Participation Benefit-Cost Ratios (Study 2)
Study 2 examined variation in benefit-cost ratios in relation to the level of Prop 36
participation. The study was based on the population of adults (18 years or older) who,
during Prop 36’s initial year (July 1, 2001 to June 30, 2002), participated in Prop 36, that is,
those who accepted a Prop 36 referral. The population was divided into three groups: (1)
offenders who were referred to Prop 36 but did not enter drug treatment, (2) offenders who
entered but did not complete treatment, and (3) offenders who completed treatment. Like
Study 1, Study 2 covered 42-month baseline and follow-up periods beginning with the date
of each offender’s conviction. Figure 7.10 provides a summary of cost offsets by treatment
status. The zero line is interpreted as cost neutral. Bars above the line represent cost
increases and bars below the line represent cost savings.
Figure 12.10
DID Cost Summary by Drug-Treatment Status

Do llars p er o ffen d er

$4,000
$2,000
$0
-$2,000
-$4,000
-$6,000
-$8,000
-$10,000
Prison

Jail

Probation

Parole

Arrest
and

No treatment

-$4,598

-$2,054

$692

-$226

$1,823

-$403

$729

-$4,037

Some treatment

-$5,694

-$1,749

$736

-$332

$2,799

$1,700

$747

-$1,792

Completed treatment -$8,425

-$1,723

$727

-$322

$1,161

$2,292

$454

-$5,836

No treatment

11

Some treatment

Treatment Health

Total

Completed treatment

Prop 36 programmatic costs are first subtracted from the numerator to avoid double counting of costs. The
benefit-cost ratio is: Ratio = ((S * N) – P)/A; where S = average savings per offender expressed as a negative
amount; N = number of Prop 36 eligibles; P = programmatic costs; A = Prop 36 allocation.

225

Prison
Drug-treatment participation was strongly associated with reductions in incarceration costs
relative to the pre-Prop 36-era group costs. Prison costs were $4,598 lower for offenders
who never entered treatment, $5,694 lower for individuals who entered but did not complete
treatment, and $8,425 lower for offenders who completed treatment, than what would have
been expected had Prop 36 not been implemented.
Jail
Jail savings were very similar for individuals who entered treatment but did not complete,
and those who completed treatment (jail costs were $1,749 and $1,723 lower, respectively).
The largest jail costs were for offenders who never entered drug treatment ($2,054 lower).
What explains the relatively greater jail savings for those offenders who never enter
treatment? UCLA found that Prop 36.-era offenders who did not report for treatment
consisted primarily of two types: offenders with low or no prior arrests and convictions and
offenders with many prior arrests and convictions. The former group may have felt they were
only recreational users not requiring treatment. The latter group may have chosen not to
participate in treatment in the belief that sanctions from the criminal justice system were too
unlikely or too distant to hold them accountable. The former group was at lower risk of rearrest and incarceration. The latter group was relatively more likely to serve a prison term
(the prison module above showed lower prison savings attributable to those who never enter
treatment).
Probation
Probation costs were $692 higher for offenders who never entered drug treatment, $736
higher for individuals who entered but did not complete treatment, and $727 higher for
offenders who completed treatment.
Parole
Parole costs were $226 lower for offenders who never entered drug treatment, $332 lower for
offenders who entered but did not complete treatment, and $322 lower for offenders who
completed treatment.
Arrests and Convictions
Arrest and conviction costs were $1,823 higher for offenders who never entered drug
treatment, $2,799 higher for offenders who entered but did not complete drug treatment, and
$1,161 higher for offenders who completed treatment.
Drug Treatment
As expected, drug treatment costs were higher depending on level of participation.
Treatment costs were $1,700 higher for offenders who entered but did not complete
treatment and $2,292 higher for offenders who completed treatment. Offenders who did not
enter treatment had a $403 lower treatment cost than similar offenders in the pre-Prop 36
period.
Healthcare
State-funded healthcare costs were $729 higher for offenders who never entered treatment,
$747 higher for offenders who entered but did not complete treatment, and $454 higher for

226

offenders who completed treatment. Our 30-month follow-up analysis in our previous cost
report showed higher healthcare costs (and therefore lower savings) for Prop 36 clients who
completed treatment. This increase indicates that offenders in treatment were more likely to
seek out care for other health needs. By 42-months this pattern of increased healthcare costs
is reversed. An analysis of quarterly healthcare costs reconciles this difference. Treatment
completers have higher healthcare costs in the short term (are more likely to seek out care),
but their healthcare costs are consistently reduced over time and ultimately result in a
healthcare cost offset (this is consistent with findings on health cost offsets for many studies
in the substance abuse treatment literature).
Total Cost Offset by Drug-Treatment Status
Total costs saved were $4,037 for offenders who were referred to Prop 36 but never entered
drug treatment, $1,792 for offenders who did not complete treatment, and $5,836 for
offenders who completed treatment. Treatment and new arrests and convictions costs
constituted a major part of cost increases, whereas total costs savings were driven largely by
savings in incarceration (jail and prison) costs.
Cost Comparison
For treatment completers, the cost savings reflect a benefit-cost ratio of about 4:1, meaning
that approximately $4 was saved under Prop 36 for every $1 allocated to a treatment
completer. Notably, although Prop 36 offenders who received some treatment showed
reductions in prison and jail time over those who did not enter treatment, these savings were
offset by treatment costs and somewhat higher rates for arrests and convictions in the followup period. Substantial savings were also found for offenders who never entered treatment. It
may be that these offenders had less serious drug problems and did not feel a need for
treatment. Further research is needed to better understand this sub-population.

Prop 36 Consecutive Cohort Study (Study 3)
Study 3 examined 30-month follow-up costs for offenders convicted during Prop 36’s first,
second, and third years. A difference-in-differences methodology was not feasible for the
Prop 36 cohort study as the estimates for first-, second-, and third-year Prop 36 offenders
were not directly comparable. The comparison group and the first-year SACPA offenders
both experienced pre-periods with no Prop 36 policy in effect. However, the second-year and
third-year Prop 36 offenders have pre-periods that extend into the Prop 36 era.
The follow-up costs are provided in Figure 11. These results suggest that Prop 36 30-month
follow-up costs for jail, probation, parole, and treatment have remained stable from year to
year. Prison costs steadily declined over the first 3 years (30-month follow-up costs in year
3 were 17% lower than follow-up costs in year 1).12 Arrest and conviction costs have
12 This is a surprising finding given data reported at the state level. In the year following the implementation
of Prop 36 there was a large reduction in admissions to state prisons for drug offenses. Since that initial decline
there has been a steady increase in admissions to state prisons for drug offenses (both due to parole revocations
and new admissions). Further research should be conducted to better understand the reduced prison-days
attributable to Prop 36 eligible offenders that we estimate from OBIS, given the increases reported in the statelevel data. It is possible that there is little overlap in these offender groups, but this should be confirmed.
Further research should be conducted to determine how much, if any, of this difference is due to those
offenders who enter Prop 36 on a parole violation, and are therefore excluded from our analysis.

227

steadily decreased (30-month follow-up costs in year 3 were 10% lower than follow-up costs
in year 1).13
Figure 12.11
30-Month Follow-up Cost Summary by Prop 36 Year

Dollars per offender

$8,000
$7,000
$6,000
$5,000
$4,000
$3,000
$2,000
$1,000
$0
Prison

Jail

Probation

Parole

Arrest and
Conviction

Treatment

Y1

$6,847

$3,701

$1,258

$563

$5,232

$1,898

Y2

$6,154

$3,547

$1,268

$566

$4,869

$1,756

Y3

$5,681

$3,757

$1,371

$534

$4,693

$1,737

Y1

Y2

Y3

Conclusions
Two major conclusions can be drawn from the benefit-cost analysis of Prop 36: (1) Prop 36
substantially reduced incarceration costs; and (2) Prop 36 resulted in greater cost savings for
some offenders than for others.
Conclusion 1: Prop 36 substantially reduced incarceration costs.
Based on costs incurred by offenders who were eligible for Prop 36 participation during its
first year of implementation, Prop 36’s overall benefit-cost ratio was nearly 2 to 1 over the
42-month follow-up period. From the state- and local-government perspectives, continued
funding of Prop 36 is justified.
Conclusion 2: Prop 36 results in greater cost savings for some eligible offenders than for
others.
In particular, drug-treatment completers had a benefit-cost ratio of 4 to 1, a savings of $5,836
per offender. Incentives should be considered for providers who demonstrate more success
in drug treatment engagement, retention, and completion for Prop 36 clients. UCLA found
that offenders with five or more convictions in the 30-month period prior to their Prop 36eligible conviction produced costs ten times higher than those of the typical offender. Prop
36 criteria should be modified so that offenders with high rates of prior non-drug convictions
13

Statewide data shows an increase in arrests for drug crimes over the same period. Further research should be
conducted to better understand the reduced arrests and conviction costs attributable to Prop 36 eligible
offenders that we estimate from administrative records. For further discussion of arrests and crime trends see
chapter 11.

228

(e.g., five or more prior convictions during the prior three years) would be placed into morecontrolled settings, including, but not limited to, residential treatment or prison- or jail-based
treatment programs.
Eligible offenders with heavy drug use should receive greater criminal justice supervision
(e.g., drug-court management or more-intensive probation or parole supervision) and more
intense drug-treatment services (e.g., residential treatment). Collaboration and coordination
among court, probation, parole, and drug treatment systems should continue to be improved
with the goal of admitting offenders into appropriate treatment in the shortest possible time,
as well as maintaining appropriate levels of oversight and supervision.
References
Ettner, S.L, Huang, D., Evans, E., Ash, D.R., Hardy, M., Jourabchi, M., & Hser, Y.
(2006). Benefit-cost in the California Treatment Outcome Project: Does substance
abuse treatment pay for itself? Health Services Research, 41, 192-213.
French, M.T. (2005). Personal communication.
Gerstein et al. (1994). Evaluating recovery services: The California Drug and Alcohol
Treatment Assessment (CALDATA). California Alcohol and Drug Programs.
Accessed at: www.adp.cahwnet.gov/pdf/caldata.pdf.
Harwood, H.J., Malhotra, D., Villarivera, C., Liu, C., Chong, U., & Gilani, J. (2002).
Cost Effectiveness and cost benefit analysis of substance abuse treatment: A
literature review. Rockville, MD: Substance Abuse and Mental Health Services
Administration. Center for Substance Abuse Treatment.
Hser et al. (2005). The California Treatment Outcome Project (CalTOP) Final Report.
UCLA Integrated Substance Abuse Programs.
http://www.uclaisap.org/caltop/FinalReport/Cover%20Page.pdf.
Miller, T., Cohen, M. & Wiersema, B. (1996). Victim Costs and Consequences: A New
Look. Washington, DC: U.S. Department of Justice, National Institute of Justice.

229

230

Chapter 13: Performance Monitoring & Outcome Measurement in
Drug Treatment Systems
Darren Urada, Ph.D. and Rachel Gonzales, Ph.D.

Calls for evidence and accountability in the substance abuse treatment field have
generated a search for appropriate measures of performance and outcomes that
stakeholders can rely upon to monitor performance.
For context, this chapter begins with a literature review to summarize the current state of
performance and outcomes measure development in the substance abuse treatment field.
Following this, specific performance and outcome measures are reviewed with the goal of
identifying the most relevant measures that, properly applied, could potentially be used to
compare county level performance and outcomes in Prop 36. Advantages, disadvantages,
and suggestions for implementation or adaptation are discussed for each.
To ensure fair cross-system and cross-program comparisons, case mix adjustments may
be necessary, potentially at both a community and client level, but the exact variables to
be used in the implementation of such an adjustment will depend on the measures to be
adjusted. Initial data from the performance and outcome measures to be adjusted will be
required to develop and optimize such adjustments. However, potential methods are
discussed and a list of variables that should be considered for adjustment is included.
Identifying information on offenders who are Prop 36 participants is required to ensure
accurate use of data from administrative databases in the future.
Further information on the services that individual clients receive would significantly
facilitate performance measurement efforts.
Logical next steps include gathering feedback from stakeholders on ideas for alternative
measures and the potential of implementing the collection of data on promising measures,
selection and development of a plan for using these measures, assessing and addressing
resource and training needs to ensure collection of high quality data, data collection, and
investigation of alternative case mix adjustment techniques where necessary.

The first step in implementing performance monitoring for treatment programs and counties
is the development of standards against which client outcomes and program performance can
be developed (McClellan, Chalk, & Bartlett, 2007). Outcomes are defined as changes in
client behaviors, and functioning that can be attributed to treatment (e.g. drug use).
Performance Measures are indicators of program use of administrative and clinical best
practices to provide quality care (e.g. treatment initiation within 14 days of the initial
request).

231

Detailed, direct, and objective measures of the quality, types, frequency (e.g., number of
counseling sessions per week), and duration of services actually delivered to individual
clients within treatment programs would be the ideal measure of performance. However, in
the absence of such direct measures, program-related performance and client-related
outcome measures described in this chapter can be used as indirect indicators to measure and
guide improvements in program performance.
Due to state interest in monitoring performance and measuring outcomes at the county level,
this chapter reviews the advantages and disadvantages of a number of performance and
outcome measures primarily focusing on measurement at that level as an example. Most of
these measures could also be applied at state or program levels, however. All performance
and outcome measures have strengths and weaknesses. Therefore, where possible, UCLA
will make suggestions on grouping specific performance and outcome measures into
complementary sets.
Once measures are identified, case-mix adjustments may need to be applied to “level the
playing field.” Individuals entering treatment can vary widely in problem severity, as can the
community context (e.g. poverty, unemployment, crime). Therefore client and county
context must be controlled for when comparing performance and outcomes across programs
or counties. Another possibility is to use measures to benchmark performance of programs
from year to year. In this instance, programs benchmark against themselves, removing the
effect of different populations between programs.
Literature Review
The goal of performance and outcome measures is to bring greater accountability to
oversight agencies and provide the basis for quality improvement and cost efficiencies by
providers of treatment services, so that better service quality and greater cost efficacy can be
achieved. Virtually all substance abuse treatment programs since the 1960s have focused on
three general client outcome goals at and immediately following treatment discharge: (1)
cessation of drug and alcohol use (abstinence); (2) reduced criminal activity; and (3)
increased productivity by way of employment (McCollister & French, 2003; Gerstein &
Lewin 1990; McClellan, 2007). McClellan and colleagues (2005) argue, however, that it is
important to take into consideration the chronic nature of dependence careers and the view of
addiction as a “health problem,” similar to diabetes or hypertension. Thus, while the
historical outcome measures should be maintained, a greater focus on health and functional
status measures appropriate for the management of a chronic illness is advised.
Under a chronic illness framework, a set of broader outcome measures should be included
that consider “health” as a state of complete physical, mental, and social well-being and not
merely the “absence of disease”. Quality in the context of healthcare is defined by the
Institute of Medicine (2001) as “the degree to which health services for individuals and
populations increase the likelihood of desired health outcomes and are consistent with
current professional knowledge”. In response, federal and state governments have begun to
establish clear criteria for what constitutes acceptable performance and outcomes
measurement. Recently, the Substance Abuse and Mental Health Services Administration
(SAMHSA) has called for the national standardization of measurement and reporting in
several areas of function as a way of evaluating all treatments they sponsor, referred to as
232

National Outcome Measures (NOMs). The goal of NOMs is to “improve service efficiency
and effectiveness through the use of indicators of program accountability and performance”
(SAMHSA, 2005). NOMs currently include the following ten domains: reduced morbidity,
employment/education, crime and criminal justice, stability in housing, social connectedness,
access/capacity, retention, perception of care, cost effectiveness, and use of evidence-based
practices (SAMHSA, 2007). NOMs is presently concentrating efforts on improving data
quality in relation to reporting in each of these 10 domains.
Performance Measurement in Alcohol and Drug Treatment
As noted above, performance measures are indicators of use of administrative and clinical
best practices. These measures differ from client outcome measures in that they capture data
on performance at the treatment programs level. They may also be applied to broader
systems such as cities, counties, states, or nations, but discussion in this chapter will focus on
programs and counties, where performance monitoring in Prop 36 is most relevant.
Performance measures have been used to draw attention to deficits and strengths in treatment
systems, identifying areas where incentives for treatment quality improvement can be
applied, and targeting areas where quality improvement is needed (Academy Health, 2004).
Performance measurement can be used to compare performance with defined
targets/benchmarks, the performance of one program or county with another, or the
performance of a program or county with its own prior performance. As such, from a policy
(funding) perspective, performance measurement can help hold these entities accountable by
identifying which are meeting or exceeding quality expectations and which are not. In
addition, performance monitoring can be very useful to treatment providers or counties by
providing feedback that can be used to improve their own services. Specifically, the
performance data derived can be used to inform decisions about what, if any, corrections
should be made to service delivery in order to improve performance. Chapter 7 in this report
addresses key aspects of implementing and assessing targeted process improvement
strategies within individual programs.
Selecting Performance Measures
Given the complexities of substance abuse and its treatment, no performance measure is
adequate to cover all facets of performance. Instead, a set of multiple complementary
measures is recommended to generate a more accurate assessment of provider performance
(McClellan et al., 2007). In 1998, the Center for Substance Abuse Treatment (CSAT)
convened the Washington Circle1, a multidisciplinary group of service providers,
researchers, managed care representatives, and public policymakers, to address the need for
performance measures for programs that treat people with alcohol and drug (AOD) disorders
(McCorry et al., 2000). This group outlined a set of performance measures that can apply to
four stages of the continuum of care:
•

1

Prevention/education: Educating patients about AOD disorders. Percentage
of adult patients with primary care visits who are advised or given
information about AOD disorders.

For more information, Washington Circle reports are available at: http://www.washingtoncircle.org/

233

•

•

•

Recognition: Number of cases per 1,000 members who were diagnosed with
AOD abuse or dependence or who received AOD-related services on an
annual basis.
Treatment: Initiation of AOD plan services within 14 days, linkage of
detoxification and AOD plan services within 14 days, treatment engagement
within 30 days of initiation of care, interventions for family
members/significant others of AOD clients in treatment.
Maintenance of treatment effects: Percentage of clients who report specific
services provided and/or monitored by the plan to promote and sustain
positive treatment outcomes after discharge.

The Washington Circle collaborated with the National Committee for Quality Assurance
(NCQA) to refine the measures. In 2004, NCQA adopted the measures for inclusion in its
Health Plan Employer Data and Information Set (HEDIS®) (NCQA 2006), which is an
information system that tracks quality of care in health plans. Two widely used Washington
circle performance measures within the HEDIS® data set that are specifically related to
addiction treatment include:
•

•

Initiation of Treatment. Initiation can be defined as the percentage of adults
diagnosed with a new episode of AOD abuse or dependence who either (1)
initiate treatment through an inpatient AOD admission or (2) have an initial
outpatient service for AOD abuse or dependence and receive any additional
AOD services within 14 days.
Treatment Engagement. This is an intermediate step between initially
accessing care (in the first visit) and completing a full course of treatment.
This measure is defined as the percentage of adults diagnosed with AOD
abuse or dependence who receive two additional AOD treatments within 30
days after initiating treatment. To qualify as a new episode, there must be a
period of 60 days, referred to as a “negative diagnosis history” or “clean”
period during which the person had no claims or encounters associated with
any diagnosis of AOD abuse or dependence.

Recently the National Quality Forum (NQF) undertook a project to address the need for
substance abuse treatment performance measures and benchmarks of effective treatment.
Measures include: screening, initiation of treatment, transition between care/coordinated
care, and medication assessment/management.2
Outcome Measurement at the Client Level
Outcomes reflect an area of life function or status at the client level that are expected to be
positively influenced by treatment. Three core sets of measures that are traditionally used in
treatment evaluations include: substance use, employment/self-support, and criminal activity.
Other indicators have included: physical health, mental health, and family or social relations
(McClellan et al. 2005). These measures are used to evaluate the state of a client’s health
resulting from the services and interventions received.
2

NQF progress reports are available at http://www.qualityforum.org/

234

Selecting Client-Level Outcomes
As discussed earlier, SAMHSA CSAT has embarked on a multi-pronged strategy to build
infrastructure and processes for performance management in all states and jurisdictions with
the mandate of NOMs collection and reporting requirements by states. Under NOMs,
consensus has been reached on ten comprehensive outcome domains to be measured and
tracked. Specific measures are being defined and standardized through cooperative efforts
involving federal and state officials, practitioners, methodologists, and researchers.
Additional activities focus on state and provider-level data reports, opportunities for
addressing accountability and program effectiveness, and strategic qualitative and
quantitative studies to facilitate performance management and advance treatment and
prevention of substance abuse.

The NOMs are client-level outcome measures that reflect public expectations:
• Increased access to services
• Increased retention in treatment
• Abstinence from alcohol and drugs
• Increased employment/education
• Decreased crime/criminal justice involvement
• Increased stability of housing
• Client social connectedness (under development)
• Client perception of care (consumer survey under development by the Forum
on Performance Measurement)
Implementation of NOMs and development of timely and flexible state data-handling and
reporting capacity will tremendously support data-based performance assessment and
management. However, this new effort will require improved and expanded data
infrastructure for many states. In recognition of these needs, all SAMHSA programs for
investment in state data infrastructure, technical assistance for improving state information
systems, and for concomitant staff training is currently being coordinated across states.
Existing Data Sources for Performance Monitoring and Outcome Measurement
Standard sources of data for performance and outcome measures include client surveys and
administrative data used to pay bills or to manage care.
Client Surveys
Surveys can capture information on how clients view their treatment experience, including
what services were received and, satisfaction with care3, satisfaction with specific treatment
components (e.g., being offered information about a medication’s side effects), and

3

Satisfaction with care: a widely used indicator of performance and quality in health care; although client
satisfaction is not well related to any other objective measure of good outcomes from treatment (e.g., urine
results or employment or re-arrest rate (McClellan & Hunkeler, 1998). In other words, while it is important
to measure a clients’ satisfaction with their treatment experience, satisfaction is not synonymous with good
outcomes (at least in the addiction field).

235

perceptions on whether outcomes of care that can be attributed to the services that they
received.
Administrative Data
In the course of providing and paying for care, insurers and related organizations typically
generate administrative data on the characteristics of the population they serve as well as the
utilization of and charges for services. While it is typical to do so at an individual-user level
for health and mental health services, reporting drug treatment services at the client level is
relatively uncommon. In most other service delivery systems, even without full
standardization of data collection, client specific data usually contain certain key elements,
such as the following:
• Date of service delivery
• Location of service
• Diagnosis and procedure codes for clinical services.
• Type of service
• Number of units (e.g., days of service)
• Amount billed and amount reimbursed, often separately by service (Garnick
et al., 2002a)

In a number of states, clients receiving AOD-related services can be identified and their
publicly or privately funded care can be tracked over time and across settings via state
administrative data. For example insurance claims data, one type of administrative data,
have been used for measuring the quality of AOD services in a variety of settings, including
commercial health plans and the Federal Government’s Medicare and Medicaid programs.
In addition, systems such as state-run specialty hospitals, Department of Veterans’ Affairs
(VA) facilities, and staff-model managed care plans deliver care through facilities they own
and providers they employ (Harris et al., 2005). Most of these systems generate encounterlevel records that have dates and descriptions of services, similar to those included in claims.
Data generated by publicly funded mental health and AOD-treatment systems administered
through State departments of public health, mental health, or substance abuse would be
similarly useful for both performance monitoring and client outcome measurement. For
client outcomes, administrative data from health, mental health, welfare, employment,
criminal justice, and other databases can be used to assess pre and post treatment client
behaviors. In California, this has been accomplished in the California Treatment Outcomes
Pilot Program (CalTOP) and in the evaluation of Prop 36. Recordkeeping of data relevant
for performance monitoring varies widely. Some states have detailed data on the types of
services provided to specific clients, but many still report only aggregate service-use
statistics or the start and end dates of episodes of care without specific detail on dates, types,
content, or duration for all services. More states are moving toward designing systems to
accumulate the level of detail required for many performance measures.
Despite the many advantages of administrative data, such data sources often have quality
problems, varying by source. For example, AOD diagnoses may not be coded accurately or

236

completely because of issues of insurability, confidentiality, stigma, or even lack of space on
the claims forms to record multiple diagnoses.
Application: County-Level Performance Monitoring and Outcome Measurement in
Prop 36:
UCLA has been asked to discuss measures that can be employed in comparing Prop 36
performance and outcomes across counties. Potential performance and outcome measures
and the advantages and disadvantages of their application are described below. In most
cases, these measures are useful for informational purposes and may be helpful for
identifying counties that potentially may need resources to enhance their performance such
as technical assistance. Ideally the measures described in this section would be used as a
first step for these purposes.
In some cases, these measures would become problematic if used in the absence of further
development as a basis for funding decisions. If tied directly to funding through some
performance algorithm, some measures could potentially create a system of perverse
incentives that may inadvertently punish positive practices and reward negative ones. Since
it is important to avoid such unintended consequences where they can be predicted, potential
problems will be noted.
All of the measures described hereafter have both strengths and weaknesses. Some are
imprecise, while others focus on only one of the many steps offenders experience in the Prop
36 process (e.g. treatment entry). Therefore, the ideal solution would be to use a set of
measures that have complementary strengths and weaknesses and together provide adequate
coverage of the Prop 36 process.
Treatment Show Rate
In Prop 36 the treatment show rate is the percentage of offenders who were referred by
courts or parole to treatment who were subsequently admitted into treatment. A similar
show rate has been reported by UCLA annually since Prop 36 was implemented based on a
combination of SRIS data, stakeholder survey data, and data from CADDS and CalOMS (see
Chapter 1). This discussion pertains to use of a simplified form of this measure, using
referral counts from SRIS and admission counts from CalOMS.
Advantages
• Treatment admissions as measured by CalOMS are relatively objective.
Counts of admissions are relatively immune to subjective bias.
• Data on treatment admissions, defined as a CalOMS treatment admission, is
collected statewide on a standard electronic measure.
• Using CalOMS to determine treatment show rates would have the added
benefit of providing an incentive to programs and counties to maximize
adherence to CalOMS admissions reporting requirements.
• The number of annual Prop 36 referrals (unique individuals) from court and
parole are already collected as part of the Prop 36 SRIS dataset.

237

Disadvantages
• While treatment admission records are readily available in CalOMS, counts
of referrals are reported by county lead agencies to the SRIS with mixed
reliability and validity. ADP, however, has undertaken substantial efforts to
verify this data in recent years.
• Exceptions may need to be made for counties that send a significant
proportion of referrals out of the county for treatment on a net basis4, since
these will not show up as in-county treatment admissions in CalOMS and
would therefore decrease the show rate. In these cases, counties should be
allowed to provide alternative counts. Likewise, based on these exceptions
ADP may need to make adjustments to reduce admission counts for counties
that have a large number of admissions from out of county referrals. ADP
may be able to measure how often this occurs by comparing the county of
client residence (using the zip code field in CalOMS) to the county that the
client received treatment in (using the county field in CalOMS).
• In counties with very few (e.g. less than 20) Prop 36 admissions per year, the
show rate may be unreliable. In these counties show rates may be
significantly affected not only by random variation in clients admitted to
treatment, but also by year-to-year differences that may cause “carry-over”
effects in small counties. That is, if a person is referred at the end of one
fiscal year but not admitted until the next, it will deflate the show rate for the
first year and inflate it for the second. In larger counties (and statewide
analyses) these carry-over effects tend to cancel out and are not a problem.
That is, people referred at the end of year 2 may not enter treatment until year
3, but as long as referrals are roughly stable from year-to-year this shortage
will be canceled out by people referred at the end of year 1 who entered
treatment in year 2. If small county funding is determined by a minimum
base allocation rather than by show rates, this will not be a problem.
• The special case of clients who opt out of Prop 36 after referral needs to be
addressed. Opting out is an offender right, and it may be unfair to penalize
counties for not placing offenders who have opted out and are no longer
required to attend treatment. Counties could be allowed to remove such
cases from their referral number. On the other hand, to the extent that optouts after referral may reflect poor information being provided to prospective
Prop 36 participants prior to their acceptance of the terms, an argument could
be made for continuing to include these offenders in the referral count for
county outcome monitoring purposes.
Issues if tied to Funding
• The intent of the measure would be to provide an incentive to counties to
improve treatment show rates, but tying funding to the show rate could
provide an incentive for counties to report lower referral numbers in an effort
to raise show rates. Although this is theoretically a number that could be
4

Clients sent out of the county for treatment minus clients received from other counties.

238

randomly audited, either through local court and parole records, the
feasibility of doing so due to resource and record access constraints would
need to be assessed. Use of statewide databases such as DOJ for audits may
be difficult due to data shortcomings in this database itself (e.g.
underreporting of Prop 36 acceptance).
Suggestions
Since the treatment placement numbers in SRIS are not currently well verified (some
counties report admissions while others report counts of unique people) UCLA recommends
using CalOMS data to count the number of unique persons admitted to treatment unless the
county can justify otherwise. Use of CalOMS data will confer certainty to the meaning of
the data and would provide an incentive to counties and providers to ensure that all
admissions are reported to CalOMS, since under-reporting would create the appearance of
lower show rates.
Conclusion
This is a promising measure, and likely the easiest promising measure to implement. ADP
currently uses a similar method to guide OTP funding.

Treatment show rate covers a very important part of the Prop 36 process, but this measure
would ideally be used along with other measures, such as treatment engagement and various
outcome measures, to provide a picture of the broader Prop 36 process.
Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Excellent
Good
Fair

Utilization of Appropriate Levels of Care
This measure refers to the appropriate use of certain levels of care based on client needs, for
example residential treatment for clients with severe drug dependence (see also Chapter 6 on
underutilization of residential treatment),
Advantages
• Use of the appropriate level of care is an important issue, both in terms of
meeting client needs (avoiding under-treatment) and cost efficiency (avoiding
over-treatment). Assistance for counties that have problems in this area
could be beneficial.
Disadvantages
• Determining appropriate care and practices requires a relevant standardized
assessment to be administered statewide and agreement on care and practices
that should be associated with the outcomes of that assessment.
Issues if tied to Funding
• Poor utilization of levels of care may often be the result of insufficient
capacity (e.g. too few residential beds) rather than assessment and placement
practices. In these cases assistance would be a more appropriate response
than a reduction of funds.

239

Suggestions
In principle this is a good idea and long term goal, but the prerequisites for placement
congruence are not yet in place. The American Society of Addiction Medicine Client
Placement Criteria (ASAM-PPC) is a commonly used tool which was specifically designed
to recommend treatment levels, but adoption of this assessment tool for Prop 36 clients is not
consistent statewide and a statewide database would ideally be created to use the results
along with actual treatment placement as a performance measure. Other measures currently
in common use statewide, such as CalOMS and ASI, were not designed for treatment
placement purposes.5
Conclusion
Overall this is an idea that has potential but additional work is required before it could have
practical application.

Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Poor
Good
Poor

Treatment Initiation within 14 Days
In the literature, treatment initiation has been defined as the percentage of adults with
substance abuse diagnoses who begin treatment within 14 days of being diagnosed as having
a need for substance abuse service (McCorry et al., 2000). Applied to Prop 36, this would
refer to a measure of the proportion of clients who began treatment within 14 days of their
referral from the court or parole.
Advantages
• Treatment initiation is relatively objective.
• Data on treatment initiation, defined by treatment admission dates in
CalOMS, is collected statewide on a standard electronic measure.
Disadvantages
• While the treatment admission date is readily available in CalOMS, dates of
referral are recorded in separate databases at the state (e.g. DOJ data) or
county level, requiring a merge of databases from separate agencies.
Identification of the referring event in these databases can be difficult or
missing depending on county level reporting (see discussion on
administrative data on arrests, convictions, violations, and incarceration). If
the referring event does not appear in these records, it is not be possible to
compute initiation within 14 days. If such missing data is random or at least
5

CalOMS contains questions related to several client characteristics at admission, some of which are related to
outcomes (e.g. residential care is associated with better outcomes among people who have used drugs more
frequently in the 30 days prior to admission), but suggesting that every client who used drugs frequently should
be placed into residential care would be an over-extension of the finding. CalOMS is not designed to be a
standardized assessment instrument. The ASI is currently the standardized assessment instrument that is closest
to being used statewide, but counties are using different versions, a few counties have resisted using it (e.g. due
to cost), and the ASI does not actually recommend certain types of treatment. Developing evidence based
standards for the types of treatment that should be administered based on ASI scores is challenging at best.

240

•

unrelated to time to treatment admission, however, cases with missing data
can generally be ignored. Existing cases would be used to compute an
aggregate proportion of cases admitted within 14 days and an assumption
would be made that the proportion is the same in the missing cases.
Acquisition of the necessary data requires collaboration among agencies
and/or researchers to obtain the data and significant time and technical
expertise are required to perform the linkage needed for this analysis.
Further work is necessary to pilot test the creation, reliability, and validity of
this measure.

Issues if tied to Funding
If residential slots are not available, the 14 day measure may encourage counties to initially
place clients in a lower level of treatment than their assessment calls for, but this is a
reasonable response and preferable to no treatment.
Suggestions
The data from DOJ is available if agency collaboration allowing access can be worked out.
Appropriate data from parole is likely available but needs to be investigated further for
feasibility. Although a minimum level of data reporting should be enforced, if initial
conviction or violation data are missing for a minority of offenders, these can be treated as
missing data without creating perverse incentives (i.e. missing data does not necessarily
make counties look “better” or “worse”). In the long run, arrangements for automatic data
sharing and linkage to use this as a performance measure would be ideal.
Conclusion
Treatment initiation within 14 days is an important and promising measure, but would
ideally be used along with other measures to provide a broader view of the Prop 36 process.
Utilization of Appropriate Level of Care would be an example of a good complement to this
measure.

Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Fair
Good
Fair

Treatment Engagement within 30 days
Treatment engagement can be defined as the percentage of clients who initiated substance
use treatment and who receive two additional visits within 30 days after the initiation of care
(McCorry et al., 2000). This differs from initiation in that it measures whether the client
“engaged” in treatment by returning after the initial visit.
Advantages
• Treatment engagement is relatively objective, assuming records of client
contacts are kept and made available.
Disadvantages
• This information is not routinely collected at the state level.

241

•

Such information may be collected in billing records at the county level,
but in many counties is likely kept for Medi-Cal clients only.

Issues if tied to Funding
• Use of this measure to determine funding would create a small incentive for
admitting clients that do not present severe problems that could prevent them
from engaging in treatment. Case-mix adjustments may at least partially
mitigate this problem, however, and, to the extent that it may be difficult to
predict engagement, this incentive may be minimized.
Suggestions
This is an objective measure with good potential if the data can be acquired.
Conclusion
Treatment engagement within 30 days is a promising measure, but new statewide
infrastructure for data collection would be required. The measure would ideally be used
along with other measures to provide a broader view of the Prop 36 process.

Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Poor
Good
Fair

90 Day Treatment Retention
Length of treatment is associated with positive outcomes, particularly if the client stays for at
least 90 days (Gerstein & Harwood, 1990; Simpson & Joe, 2004).
Advantages
• 90-day treatment retention rates can be calculated from CalOMS
admission and discharge dates, which are collected on a standard
electronic measure statewide.
Disadvantages
• This measure implies a one-size-fits-all approach. Not all clients may need
90 days, while for others 90 days will be insufficient.
• A proportion of discharge records are often missing in databases of this type.
The extent of missing discharge information will need to be assessed in
CalOMS as treatment programs and counties adjust to this new database.
• This measure does not take into account the types, intensity, or frequency of
services provided during the 90 days, which may vary widely by program
and county.
• ADP provides a specific definition for discharge dates and specifies the
conditions under which an administrative discharge should be carried out
“when a person stops appearing for treatment without notice” (ADP, 2007).
Program adherence to these instructions should be assessed before tying
funding to such a measure. If discharge dates are not being reported
according to a statewide standard, variations will be difficult to interpret and
the usefulness of this measure will be undermined.

242

Issues if tied to Funding
• Use of this measure to determine funding would create an incentive for
admitting clients with better prognoses while discouraging admissions of
clients with less positive prognoses. Case-mix adjustments, however, may at
least partially mitigate this problem.
• Use of this measure would provide an incentive to budget-strapped counties
to decrease service intensity in order to increase length of stay. Such a
decrease in intensity would be an unintended negative consequence of using
this measure.
Suggestions
To mitigate the incentive for admitting only clients with positive prognoses, this measure
could be calculated at the county level in Prop 36 using the county number of offenders
referred to treatment:

# of clients in the county retained for 90 days / # referred to treatment
Conclusions
Treatment retention is an important issue, and is useful for informational purposes. Counties
that are having difficulties on this measure might be good candidates for technical assistance
and training. However, this measure would be best used in a package linked with a measure
that assesses the services provided during the 90 day period such as use of evidence-based
practices. Otherwise, given the disadvantages listed above, the measure may not be a good
candidate for county outcome monitoring linked to funding.

Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Good
Good
Poor

Use of Evidence Based Practices
Advantages
• Use of evidence based practices is the most promising way to improve
treatment.
• Some researchers (Friedmann, Taxman, & Henderson, 2007) have
successfully used survey instruments to measure use of evidence based
practices.
Disadvantages
• Not all practices may be adequately measured by surveys. Other types of
measures known as fidelity scales have been developed to ensure that
treatment is delivered as specified by certain standards for a few evidence
based practices (e.g. Bond et al., 2000). However fidelity measures are very
resource intensive. Such measures typically involve full-day, in-person
assessments including interviews, observations, and chart reviews. Further
research is required to determine what measures can be practically deployed
for the purpose of statewide monitoring.

243

Issues if tied to Funding
• Measures of the use of evidence based practices to guide funding could create
an incentive to report that these practices are in use.
Suggestions
Assessment via fidelity scales would avoid the problems of self-report measures, but this
advantage must be weighed carefully against the resource-intensive nature of these scales.
Conclusions
This is a very important issue, and no other measure may be as important for treatment
improvement, but further work is required before this measure will be ready to deploy. If
valid and reliable data can be collected, they will be useful. The primary barrier is the
collection of such data.

Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Poor
Good
Fair

Treatment Experience/Treatment Satisfaction
Advantages
• Client treatment satisfaction is an important outcome on its own,
independent of its relationship to other outcomes.
Disadvantages
• These measures necessarily only cover clients who are admitted to treatment.
• These measures would require the collection of new survey data statewide.
Alternatively, a set of questions could be added to the discharge questions on
CalOMS, but this would have the disadvantage of only being collected in exit
interviews (see CalOMS outcomes section for disadvantages associated with
this). Furthermore, CalOMS exit interviews are not anonymous and would
generally be conducted by the treatment provider being rated, so clients may
not feel free to provide honest answers.
• Treatment satisfaction is not consistently associated with improved outcomes
on other measures.
Issues if tied to Funding
• Tying satisfaction to funding would create pressure on providers to obtain
better ratings. Incentives for methods of increasing ratings via means other
than actually increasing client satisfaction can easily be prevented, however
(see suggestions).
Suggestions
Data should be collected consistently across providers and counties using a standardized set
of questions. Ideally, the survey would be conducted by a 3rd party with anonymity assured
to the client. However, given a lack of consistent association with other outcomes, treatment
satisfaction would be most useful within a package of other measures.

244

Conclusions
This is an important issue, and may be particularly informative for improving treatment at
individual treatment providers, but given the disadvantages listed above, the measure is not
currently a good candidate for comparing counties for funding purposes.

Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Poor
Fair
Poor

Treatment Discharge Status: Completion
Prop 36 treatment completion has become the focus of much discussion. Data on whether a
participant has completed treatment is collected as part of every discharge record in
CalOMS, as it was in its predecessor, CADDS. CalOMS defines treatment completion as
occurring “when a program participant completes his/her treatment/recovery plan.” The
discharge status also indicates whether the participant has been referred for further services
or not.
In Prop 36, treatment completion is defined further. Penal Code 1210 states:
The term "successful completion of treatment" means that a defendant who has
had drug treatment imposed as a condition of probation has completed the
prescribed course of drug treatment as recommended by the treatment provider
and ordered by the court and, as a result, there is reasonable cause to believe that
the defendant will not abuse controlled substances in the future. Completion of
treatment shall not require cessation of narcotic replacement therapy.
Advantages
• Discharge status is recorded on a standard electronic measure statewide.
• Discharge status data is reasonably complete.
• Discharge status data is readily available from a single statewide database.
• If the definition of completion is standard and stable, programs within a
similar modality can be compared to others or to their own record from
year to year.
• This measure can give policymakers an idea of how often clients complete
their treatment plans, despite the imprecise nature of the measure.
• Treatment completion is part of an offender’s requirements for fulfilling
his/her obligations under Prop 36.
Disadvantages
• Clients’ treatment/recovery plans will differ widely between different types
of treatment. Requirements and plans can and should be different between
detoxification, outpatient, intensive outpatient, residential, and methadone
maintenance programs. Plans may also differ from client to client within a
treatment program. Therefore, “completion” can mean entirely different
things in different contexts. Also CalOMS now includes “Complete but

245

•
•

referred” separate from “Complete and not referred”, further underscoring the
variability in the meaning of completion.
Even where treatment plans are the same, the determination of whether a
client has completed those plans may be subjective.
Some providers may resist use of the discharge status “complete” on the
basis of its perceived inconsistency with a chronic illness model of drug
dependence.

Issues if tied to Funding
• If linked directly to funding, this measure would create a perverse incentive
for counties to encourage programs to redefine completion requirements in
ways that will make higher completion rates easier to achieve (e.g. reducing
treatment plan goals).6 This may have an adverse impact on clients. The
existence of such an incentive does not mean most counties would respond in
such a way, but it is critical to understand that because counties are
essentially in a zero-sum competition for Prop 36 budget dollars, the actions
of any one county that result in increased funding impacts all other counties.
Therefore even if nearly all counties resist loosening completion
requirements, they could unfairly lose a share of funding if any other county
does so. If another county follows suit to avoid loss of funding, pressure on
remaining counties would continue to escalate.
• Treatment completion only applies to offenders who entered treatment.
Therefore using treatment completion to determine funding would create an
incentive system that could reward high treatment no-show rates if selfselection of clients results in the most motivated clients making it to
treatment. Put differently, higher no-shows may tend to remove clients that
have a lower chance of completing treatment, therefore the system may
reward high no-show rates if treatment completion is considered in isolation.
Suggestions
If stakeholders wish to use treatment completion as one of a set of variables to compare
outcomes among counties or providers, a much more specific, standard definition of
completion would need to be created to ensure that completion means the same thing in each
program. Some counties have initiated efforts to standardize definitions within their
counties, so it would make sense to examine whether any of these efforts can be applied
statewide. However, standardization may necessarily involve restrictions on treatment
program flexibility in defining individual treatment plans, which is a major consequence that
must be weighed carefully against the advantages of standardization.

One way to try to mitigate potential incentives to reduce show rates by admitting only those
most likely to complete may be addressed in Prop 36 by instead using the percentage of
6

Even if funding is linked to completion on the county level (i.e. county funding based on aggregated
countywide completion rates), this does not substantially mitigate the problem because programs are usually
either run by their county directly, or else are under contract with the county and may have their definitions of
completion influenced either directly (e.g. via county policies on allowable length of treatment), or indirectly
via knowledge of the effect of completion rates on funding.

246

offenders who were initially referred to treatment who completed treatment as opposed to the
percentage of clients admitted to treatment who completed. However, this is not
straightforward since clients can opt out of Prop 36 after being referred. Opting out is within
participant rights, and it may be unfair to penalize counties for offenders who did not
complete treatment because they have opted out and are no longer required to attend
treatment. Counties could be allowed to remove this from their referral number. On the
other hand, to the extent that opt-outs after referral may reflect poor information being
provided to prospective Prop 36 participants prior to their acceptance of the terms, it is not
completely out of the control of county stakeholders. Using both methods of calculation
may provide a range score that may be a comparison alternative.
Completion of detoxification should not normally be considered “treatment completion”,
since it is considered to be a “pre-treatment” step.
Conclusions
The use of treatment completion as a single measure in isolation for determining funding in
Prop 36 is not recommended due to the potential perverse incentives that this could create.

Treatment completion is an imperfect measure but could be useful as an informational tool to
make comparisons in which the completion definition does not change (e.g. the same county
over time assuming the mix of service modalities, clients, and treatment completion policies
in the county remains relatively stable). Treatment completion could also be useful as the
first step in assessing whether some counties may need technical assistance.
Due to its weaknesses, ideally, treatment completion is better used in conjunction with other
measures as part of a package that together provides a more complete look at what happens
before treatment admission (e.g. show rates), services provided during treatment, and followup measures after treatment discharge.
Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Excellent
Fair
Poor

Treatment Discharge Status: Completion plus Satisfactory Progress
Data on whether a participant made satisfactory progress when treatment was not completed
is collected as part of every discharge record in CalOMS, as it was in its predecessor,
CADDS. CalOMS defines satisfactory progress simply as occurring “when a participant has
made satisfactory progress in a program.” The data also indicate whether the participant has
been referred for further services or not. It is possible to combine this discharge status with
treatment completion status to generate an indicator for clients that either completed
treatment or made satisfactory progress.
Advantages
• Discharge status is collected on a standard electronic measure statewide.
• Discharge status data is reasonably complete.
• Discharge status data is readily available from a single statewide database.

247

Disadvantages
• Since treatment completion is a part of this measure, it suffers from all of the
same disadvantages listed above for treatment completion.
• Treatment completion is required to fulfill the obligations of Prop 36.
Incomplete treatment with satisfactory progress does not fulfill this
requirement.
• “Satisfactory Progress” is even more subjective than treatment completion.
Unlike treatment completion, which is at least tied to progress on a treatment
plan, satisfactory progress is not defined statewide in any detail.
Issues if tied to Funding
• As with treatment completion, if linked to funding, this measure could create
an incentive for programs to redefine satisfactory progress in ways that will
make higher rates easier to achieve. Given the subjective nature of the
measure, tying the measure to funding would encourage greater use of the
“satisfactory” discharge status even if no real changes in client outcomes or
treatment policies occur. The existence of such an incentive does not mean
most counties would respond in such a way, but because counties are
essentially in a zero-sum competition for Prop 36 budget dollars, the actions
of any one county can impact funding for others. Therefore even if the
majority of counties resist loosening the definition of satisfactory progress,
counties could unfairly lose a share of funding if any other county does so.
• Satisfactory progress only applies to offenders who entered treatment.
Therefore using treatment completion plus satisfactory progress as a single
measure to determine funding would create an incentive system that rewards
high treatment no-show rates if self-selection of clients results in the most
motivated clients making it to treatment.
Suggestions
In order to be made useful, first the definition of treatment completion would need to be
standardized statewide as described in the section on completion. Following this, a standard
definition of “satisfactory progress” would need to be agreed upon and enforced.
Conclusions
Treatment completion plus satisfactory progress is a very questionable measure in the
context of Prop 36, where completion is required. It could be somewhat useful as a purely
informational tool to make comparisons in which the definition is stable (e.g. the same
county over time, assuming the mix of service modalities, clients, and practices for
measuring completion and satisfactory progress remains relatively stable), and as the first
step in assessing whether some counties may need assistance. However, due to the
subjective nature of the measure it is currently difficult to interpret on its own even in this
context.

Due to its weaknesses, treatment completion plus satisfactory progress could be used in
conjunction with other measures as part of a package that together provides a fuller look at
what happens before treatment admission (e.g. show rates), services provided during
248

treatment, and follow-up measures after treatment discharge. But absent this context it is not
recommended.
Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Excellent
Poor
Poor

CalOMS Outcome Measures
CalOMS contains several outcome variables, which are measured upon admission and at
discharge. It is possible to compute changes on these measures from admission to
discharge. These fall into several domains:
• Drug use
• Employment
• Criminal justice
• Medical/Physical health
• Mental health
• Family/Social
Advantages
• CalOMS outcome variables are recorded on a standard electronic measure
statewide.
• CalOMS outcome variables are readily available from a single statewide
database.
• These measures reflect dimensions that are widely recognized as important
outcome variables, as evidenced by their inclusion in NOMs.
• The CalOMS outcome variables allow comparisons of measures for the same
client from admission to discharge. This pre-post design mitigates some of
the effect of differences in client characteristics between counties.
Disadvantages
• CalOMS is a very new data system, which started collecting data January 1,
2006, and system evaluation efforts have just begun. More needs to be
known about data quality and reporting practices. Education and training for
providers reporting data to CalOMS may be required before this data is used
for county comparison purposes.
• CalOMS only covers the subpopulation of Prop 36 offenders who are
admitted to treatment.
• Only clients who were discharged with an exit interview have discharge
outcome data. Losing information of those clients who simply drop out of
treatment could potentially produce selection bias since treatment completers
may be more likely to complete an exit interview than unsuccessful clients,
who sometimes stop attending treatment without warning.

249

Issues if tied to Funding
• Because this is a very new data system, issues related to data quality are not
yet understood.
• Because CalOMS only covers offenders who are admitted to treatment, this
could confer an unintended advantage to counties with low show rates if only
the most motivated offenders show up for treatment. This situation could
create a small incentive for admitting clients who appear to be the most
motivated, while creating a disincentive to admit those with poorer
prognoses. However this problem is partially offset by the fact that client
outcomes would be tracked from admission to discharge, clients with more
negative measures at admission actually have more “room to improve” so the
direction of the incentive is not entirely clear. Furthermore, proper case-mix
adjustment can further mitigate this issue.
Suggestions
Since the CalOMS outcome variables only cover people who were admitted to treatment,
these variables should be used in conjunction with complementary measures (e.g. show rates,
treatment initiation) that can take into account the proportion of clients that are not admitted
to treatment.

Although it is theoretically possible to statistically estimate (impute) outcomes that are
missing due to administrative discharges, the precision of such estimates cannot be
guaranteed, so such analyses should be limited to informational purposes only, and would be
inappropriate for making county by county comparisons.
Implementation of post-discharge interviews would provide a favorable alternative by
collecting information on clients who had administrative discharges in addition to clients
discharged with exit interviews. If implemented, such a follow-up would mitigate the
problem associated with administrative discharges, but it would not solve the problem of noshows who never enter treatment as required under Prop 36. Therefore an adjustment for noshows as described above would still be necessary.
Further research on CalOMS will be required to improve understanding of the data and data
collection practices associated with this dataset and examine options for using this data.
UCLA is currently conducting a CalOMS evaluation that may inform these efforts.
Conclusion
The CalOMS outcome measures provide rich data on highly relevant topics. These measures
are useful for informational purposes to guide county needs (e.g., if employment outcomes
are low, this may suggest further assistance or resources in this area are needed). However,
several data issues are problematic and further development and study is necessary before
these outcomes can be deployed as a set of measures to be used as a guide for funding
decisions. Further work is necessary to better understand data collection practices and
identify optimal ways to use this data.

Since CalOMS only covers offenders who were admitted to treatment, CalOMS outcome
measures would be best used within a package of Prop 36 measures that provide
250

representation for offenders who do not make it to treatment (e.g. treatment initiation, show
rates). Since CalOMS outcome variables are based on self-report data, it would be helpful to
assess the validity of these responses by linking client information to data in administrative
databases and conducting reliability and validity studies.
Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Excellent
Fair
Poor

Prop 36 Completion
Fully completing the Prop 36 requires completion of all terms of probation or parole in
addition to treatment completion.
Advantages
• Represents the most comprehensive definition of “completion” as defined
and intended under Prop 36.
Disadvantages
• Drug treatment completion is a requirement for Prop 36 completion. This
means Prop 36 completion suffers from all of the disadvantages associated
with the ambiguous and subjective definition of treatment completion.
• In addition, this measure suffers from wide variation in the additional
requirements that an offender must complete, which may vary county by
county.
• Prop 36 completion may be underreported in statewide DOJ databases, where
it exists as disposition information. Further investigation is warranted prior
to use of this measure.
• Prop 36 completion takes a relatively long time. Some counties require as
much as 3 years on probation before an offender fully completes. Therefore,
current completers reflect the result of efforts and policies that may have
been in place years ago. This undermines the measure’s usefulness as an
assessment of recent practices.
Issues if tied to Funding
• If linked to funding, Prop 36 completion rates could create a perverse
incentive for counties to “lower the bar” by redefining completion in ways
that will make higher rates easier to achieve. For example, there could be an
incentive for counties that require an offender to be employed or enrolled in
school to drop this requirement, or it may provide an incentive for reducing
the length of time offenders must spend under supervision and drug testing
before they complete. Under such changes completion rates may rise but real
outcomes may worsen since resources to clients may actually decline. The
existence of such an incentive does not mean most counties would typically
respond in such a way, but because counties are essentially in a zero-sum
competition for Prop 36 budget dollars, the actions of any one county can
impact funding for others. Therefore even if most counties resist loosening

251

the definition of completion, these counties could unfairly lose a share of
funding if any other county does so.
Suggestions
Specific requirements for completion could be imposed statewide, but the significant delays
and data problems associated with this measure make it a generally poor candidate for
comparing counties.
Conclusions
While it has some informational value, Prop 36 completion is, at present, not an appropriate
measure for outcome monitoring on its own. Even if very specific requirements for
completion were imposed statewide, the significant delays and data problems associated with
this measure make it a poor candidate.

Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Excellent
Poor
Poor

Treatment Re-Entry
Stakeholders have expressed some interest in tracking treatment re-entry, which can be
tracked in CalOMS in the form of admissions after an initial discharge. This measure
presents a challenge as an outcome measure, however, since re-entry can be seen either as a
failure of prior treatment or as a success indicating the client recognizes treatment benefits
and may have appropriately returned before a lapse became a full relapse.
Advantages
• Data is collected on a standard electronic measure statewide.
• Data is reasonably complete.
• Data is readily available from a single statewide database.
• Re-admissions are important from a chronic care standpoint.
• Collecting information on re-entry is useful to understand other outcomes
(i.e., continued drug use, crime levels, etc). At the client level, multiple readmissions may indicate a need for a different level of care than the client has
been receiving.
Disadvantages
• This measure is difficult to interpret specifically for county outcome
comparison purposes since it represents an incomplete picture without further
information.
Issues if tied to Funding
Although treatment re-admissions can be a positive event in some circumstances, care must
be taken to avoid simply providing an incentive for increasing re-admissions, since this could
encourage lower treatment retention (which would increase the pool of clients for readmission).

252

Suggestions
This is an important measure, but because it is not necessarily a negative or positive outcome
it becomes problematic in the context of comparing counties. Even in combination with
variables such as treatment completion (from the initial treatment), re-entry suffers from
significant ambiguity specifically as a county outcome variable.
Conclusions
Although this measure has informational value and may especially be useful at the client
level, treatment re-entry is not an appropriate measure for the purpose of comparing county
performance and outcomes except as contextual background.

Current data availability:
Excellent
Potential for county-level informational use:
Ambiguous
Potential for use in comparing counties for funding purposes:
Poor
Administrative Data on Arrests, Convictions, Violations, and Incarceration
Data on arrests and convictions are collected by county sources and sent to DOJ. Data on
violations and re-incarceration are maintained by CDCR. Data on probation violations is
available at the county level.
Advantages
• These measures are of great interest due to public safety and cost
concerns.
Disadvantages
• Arrests, convictions, violations, and incarceration depend not only on
criminal behavior, but also on law enforcement practices, discretion on the
part of multiple enforcement, supervisory, judicial, and legal actors, and the
effectiveness of prosecution efforts. Outcomes will appear “worse” in
counties where enforcement is more active (arrests, probation/parole
violations) or where prosecution is more effective (convictions).
• Reporting of arrests and convictions is somewhat uneven. An unknown
number of arrests and convictions are not reported to DOJ, and
underreporting bias is likely to be associated with reporting problems and
practices at the county level. Therefore if lower arrest or conviction rates are
tied to funding, this would have the unintended consequence of rewarding
counties that are poorer at reporting these events to DOJ.
• Counting arrests requires the allowance of an adequate time period to allow
arrests to occur, plus time to allow for reporting delays. This undermines the
measure’s usefulness as an assessment of recent practices.
• In addition to the delays associated with analyzing arrests, many months
often pass between an arrest and a resulting conviction. Therefore
convictions are a lagging indicator of policies and practices. This further
undermines the measure’s usefulness as an assessment of recent practices.

253

•

•

Data on probation violations in particular are often not reported to DOJ.
Therefore this data would need to be collected at the county level, which
would require a major effort to standardize data statewide.
In any case, these analyses require positively identifying Prop 36 participants
within these relevant databases. This is not a simple endeavor, and is
currently not possible to do uniformly well at the county level using a
statewide database. The identification of Prop 36 participants in statewide
databases is incomplete. For example, a disposition indicating participation
is available as an option in DOJ data but is not always used. In these cases
participants can only be identified via linkage to other databases that indicate
participation, such as data provided by counties or CADDS/CalOMS data.
However, CADDS/CalOMS data only provide information on participants
who actually entered treatment.

Issues if tied to Funding
• If these measures were to be used as an outcome to determine funding on a
county by county level it would effectively penalize counties that have
effective enforcement and reward counties with relatively lax enforcement.
For example if arrests based on bench warrants are counted, arrests will be
higher in counties that actively pursue offenders with outstanding warrants.
Some agencies actively search for such offenders while others generally
make an arrest based on the warrant only if the offender is encountered in the
course of other law enforcement activities. Therefore, if lower arrest rates
are linked to funding, this will actually punish counties that search for
offenders with outstanding warrants and reward counties that are relatively
lax in enforcement. This would run counter to UCLA’s recommendations on
swift and certain sanctions (see Chapter 10).7
Suggestions
Conviction and violation data have all of the problems inherent in arrest data plus additional
ones. Arrests, while not ideal, come “closer to the crime” than other data in criminal justice
records and are most commonly used by criminologists to measure re-offending (Maltz,
2001). If arrests are to be used to compare individual counties, however, it is necessary to
understand the limitations inherent in these data. Coordination with DOJ would be important
to understand variations in county data.

The simplest and most promising way to take county differences into account may be to use
pre-post measures, for example per-offender arrests during the 12 months prior to Prop 36
entry and arrests 12 months following Prop 36 entry. While counties have different law
enforcement and data reporting practices, as long as these practices remain fairly consistent
7

The fact that arrests are generally carried out by agencies other than the agency that will be affected by
funding may somewhat mitigate the likelihood of unintended consequences. That is, law enforcement agencies
may not actually change their arresting behavior based on concerns over whether this will affect Prop 36
funding being funneled into the county’s lead agency. However, even if the potential perverse incentives do
not actually result in negative policy changes, the issue of whether it is fair for counties with stronger law
enforcement to receive less funding remains a significant issue.

254

over time they can be expected to roughly cancel each other out using the pre-post method.
For example, if County A and County B have the same re-arrest rate but there is consistently
0% under-reporting in county A and 20% under-reporting in county B, the data will
erroneously show County B has a 20% lower re-arrest rate in the follow-up period.
However, if instead the difference between County B’s arrest rates over the 12 months prior
and the 12 months after offenders enter Prop 36 is used, County B will have 20% lower rates
in both periods so County A and County B should have the same pre-post arrest outcomes.
Similarly this can also partly control for other differences in counties such as law
enforcement practices and client characteristics as long as these are stable over time.
Additional efforts to statistically control for these factors and address other issues would still
be advisable, however. For example, bench warrant arrests should not be included as rearrests for county comparison purposes for two reasons: First, these are typically the result of
an offender failing to appear at a Prop 36 court hearing. Arrests based on Prop 36 failures to
appear would not be adequately controlled for in the pre-post methods because they do not
meet the requirement of being stable over time (they occur only in the post-period, after the
person has entered Prop 36).8 Therefore, they would introduce county-level bias (based on
county practices and reporting) into the measure. Second, counties should be rewarded
rather than penalized for quickly finding offenders who fail to appear and bringing them
back into the system.
It would be extremely helpful to collect identifying information on Prop 36 participants in all
58 counties based on data from county sources, but this has not been accomplished to date.9
The smallest possible sufficient dataset would consist of only two variables: offenders’
criminal investigation and identification (CII) number (assigned by DOJ) and the date of the
Prop 36 referral (conviction or violation). From there, analysts with sufficient knowledge
and permission to access DOJ and CDCR data can retrieve the records needed. In the
absence of CII information, other identifiers (e.g., name, date of birth, sex, social security
number) would be required to identify the correct individuals in these databases.
Conclusions
These measures can be informative for general purposes on a statewide basis, but several
factors inhibit them from being useful as county by county outcome measures tied to
funding.

The following ratings represent use of any of these measures for re-arrest (follow-up only)
reporting given current data.
Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:
8

Fair/Poor
Poor
Poor

The exception to this case is offenders who had an existing Prop 36 case and picked up a new case. In these
instances, they would be subject to all Prop 36 arrest practices in both the pre- and post- period.
9
For the initial Prop 36 evaluation, UCLA collected this information from 10 counties. This required
individual agreements with each county, resulted in 10 different sets of data with differing formats and
definitions, and the flow of data ended along with the evaluation. A continuing statewide effort to collect
standard data from all 58 counties may be best led by ADP.

255

The following ratings apply to use of pre-post arrest data, if identifying information and
dates on Prop 36 participants were available from the counties in a statewide data set:
Current data availability:
Potential for informational use:
Potential for use in comparing counties for funding purposes:

Good
Good
Fair

Other Statewide Administrative Data: Health, Mental Health, Employment,
Welfare, Child Welfare
Aside from previously described criminal justice and treatment data, data on a variety of
other domains are also available, including health (Department of Health Care Services),
mental health (Department of Mental Health), employment (Employment Development
Department), welfare (Department of Social Services), and Child Welfare (Department of
Social Services).
Advantages
• These measures are available in statewide databases.
• These measures cover areas that can substantially impact client well being
as well as taxpayer costs.
• These are relatively objective measures.
Disadvantages
• Analyses of these databases require positively identifying Prop 36
participants within these relevant databases. This is not a simple endeavor,
and is currently not possible to do uniformly well at the county level using
statewide databases. See disadvantages of Administrative Data on Arrests,
Convictions, Violations, and Incarceration for further discussion of these
issues.
• Not all outcomes are necessarily positive or negative. For example, health
and mental health service utilization may rise, not necessarily because Prop
36 clients are having more health problems, but more likely because clients
are beginning to take care of problems that went untreated while they were
using drugs. In that context, increased utilization can be seen as a positive
outcome. Similarly, employment may drop and welfare costs may rise
initially as the client connects with social services and attends to treatment
obligations. Therefore while this data is very informative for some purposes,
interpretation of the data to compare counties would need to be done with
care.
• Data sharing practices vary between state agencies. It is fairly difficult to
obtain data from certain agencies due to wide variation in agency policies.
For example, EDD did not provide ADP with requested EDD data for this
evaluation in 2007.
Issues if tied to Funding
• Since not all outcomes are necessarily positive or negative, analyses meant to
be used for comparing counties must be chosen carefully comparisons.

256

Suggestions
Limited comparisons could be conducted, for example, changes in employment during the
fiscal quarter after treatment discharge could be compared to employment during the fiscal
quarter before discharge.
Conclusions
These measures can be informative for general purposes on a statewide basis, but they have
limited use in the specific context of county by county comparisons.

Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Fair/Poor
Poor
Poor

The following ratings apply to use of data with identifying information and dates on Prop 36
participants available from the counties:
Current data availability:
Potential for informational use:
Potential for use in comparing counties for funding purposes:

Good
Good
Fair

Drug Testing
Advantages
• The tests themselves (as opposed to testing practices) are fairly standard and
highly reliable.
• Testing has generally already been implemented statewide, though
practices and policies vary.
• Positive test results identify clients who may need more intensive services
or greater supervision.
Disadvantages
• Currently no statewide database on drug test results exists. Therefore a
database and infrastructure for collecting this data would need to be built.
This may be fairly difficult because tests are conducted by various entities
(treatment, probation, and parole).
• Drug testing measures drug use only, which is important but is only a part of
the outcomes picture. Relapse is considered to be a normal part of the
recovery process and is expected within a chronic disease model of addiction.
Therefore, while drug testing can be useful as part of a broader package of
measures it may be insufficient as a stand alone measure.
• Drug testing as a measure for comparing county by county outcomes would
require standardization of testing practices, which will necessarily reduce the
testing flexibility currently available to each county. Standardization will be
necessary for many testing protocols (e.g. testing should be randomly
scheduled), but further work is necessary to see if standardization of
frequency is necessary. In practice, reducing the frequency of drug testing is

257

•

•

often used as an incentive to reward clients that are doing well, and research
generally supports the use of incentives (contingency management) in
conjunction with treatment. Therefore ideal methods for statistically
adjusting for differences in testing frequency across counties and individuals
would be developed, thereby maintaining the ability to vary testing
frequency.
If statistical controls for differences in drug testing frequency across counties
and individuals cannot be successfully developed, the alternative, requiring
all counties to test equally, may lead to further unintended consequences.
Due to budgetary constraints the agreed-upon testing standards would likely
be less stringent (e.g. less frequent) than practices currently in place in some
counties. Therefore the consequence of implementing identical standard drug
testing procedures statewide would be the weakening of testing procedures in
some counties.
In any case some minimum level of testing will need to be in place, which
will essentially create a minimum budget requirement for all counties to
maintain testing at the specified level. Prop 36 funds cannot be used for
testing, and Substance Abuse Treatment and Testing Accountability funds
may not be sufficient depending on level of testing agreed upon, which could
create an under-funded mandate in a worst-case scenario. If this happens
counties may need to bear the cost of some testing.

Issues if tied to Funding
• Using this measure to compare counties while failing to effectively standardize
testing practices would actually penalize counties that adhere to testing best practices
(i.e. counties that have random and frequent tests will have more “dirty” tests than
counties that have predictably scheduled and infrequent tests).
Suggestions
If a decision is made to pursue this measure, a comprehensive assessment of drug testing
policies and practices across the state would be a useful first step. Where policies and
practices do not match best practices, technical assistance could be offered. In the meantime,
the feasibility of building a statewide database could be assessed. Difficulties with this effort
may be compounded by the current environment of unstable funding, which is not conducive
to long-term planning. If these hurdles can be surmounted, plans for statistically controlling
for variations in test frequencies and other factors based on the assessment of policies and
practices should be developed followed by possible standardization of practices (contingent
on the assessment of practices and whether development of statistical techniques was
successful) and ultimately data collection.
Conclusion
In theory, drug testing has potential, but serious practical issues remain. If appropriate
statistical and practical issues can be resolved, this should be regarded as a promising
measure, but if not this measure would not be recommended.

Since drug test results provide only part of the outcomes picture, they would be best used
within a package of measures that include other outcomes and treatment process measures.
258

Current data availability:
Potential for county-level informational use:
Potential for use in comparing counties for funding purposes:

Poor
Good
Fair

Case-Mix Adjustment
Two treatment providers may serve client populations that are markedly different in
demographic characteristics, drug use patterns, criminal histories, and a myriad of other
background variables. Therefore it would be unfair to compare performance and outcomes
in these two providers without making a statistical allowance for these differences. This is
known as a case-mix adjustment.
Case-mix adjustments can also be carried out at the community or county level. As with
providers, treatment clients served in one area or county may differ from those served in
another. In addition, counties may also differ in provider-level characteristics. For example
some counties make greater use of residential treatment than others, which can have an
impact on measures such as length of stay and treatment completion rates, which can
systematically differ by service type.
After case-mix adjustment is applied, performance or outcome measures can be either
compared to a standard benchmark or used to create a ranking system among peers. Counties
can be assessed against a standard benchmark or percentile rankings can be generated to
allow a county to see where it stands in comparison to other counties with a statistically
standardized case mix. For example, all counties can be ranked on 14 day treatment
initiation to know if they are in the highest 10% of counties on this measure. Counties on the
high end of the rankings could be recognized and rewarded, while those at the low end of the
rankings could be contacted to assess what types of assistance (e.g. training) may be needed.
The general approach to analysis using case-mix adjustment strategies is as follows:
•
•

•
•
•

Select a set of outcome or performance measures of interest to be used to
compare counties
Select a set of client or provider level characteristics known to predict the
outcome or performance measures of interest. These variables measure the
“case-mix” for which the adjustment is being made.
Statistically generate predicted outcomes for individual clients and sum for
each county
Compare aggregated actual outcomes to predicted outcomes for each county
Test for statistical differences. Tests can indicate differences between
expected and actual outcomes, differences between individual counties and
the median county, or differences between individual counties and a known
top performing county.

Different multivariate statistical models can be used to generate the predicted outcomes.
Koenig and colleagues (2000) used three different types of regression analyses to perform
case-mix adjustments on the same outcome variable (employment) among substance abuse
treatment providers and found that estimates of rankings varied little across the three models.

259

However, the model that can be used will be dictated by the characteristics of the measures
being analyzed (e.g. if the data is categorical or continuous).
Selection of Predictor Variables
Selection of the variables that should be used in a case-mix adjustment depends in part on
which measures are being used to rank counties. Different variables can be expected to
predict different outcomes.

For example, often the measure of an outcome in the period prior to treatment is the
strongest predictor of the same outcome following treatment. Among methamphetamine
users, Hillhouse and colleagues (2007) found that pre-treatment methamphetamine use was
the most consistent predictor of in-treatment performance and post-treatment outcomes
(gender, route of administration and pre-treatment methamphetamine use were also
significant predictors). Similarly, McCamant and colleagues (2007) found that substance use
at 1-year follow-up was strongly predicted by measures of substance use prior to treatment (a
combination of age at first use, last regular use, and frequency of use at admission).
Still, certain measures are commonly tested and found to be associated with a wide variety of
treatment outcomes, such as education, employment, drug use severity, and mental illness.
For example:
Butzin and colleagues (2002) found that among participants in a drug court diversion
program, those who were most likely to successfully complete treatment were at least high
school educated, employed, and used drugs less frequently.
Brecht and colleagues (2005) found that, among methamphetamine users, those who had at
least a high school education were older at treatment admission, did not have a disability,
had lower severity of methamphetamine use, and were not using injection drugs were more
likely to complete treatment and had longer treatment retention.
Hiller and colleagues (1999) found that early treatment dropout was related to cocaine
dependence, having a history of psychiatric treatment, being unemployed before adjudication
to treatment, and higher levels of depression, anxiety, and hostility at intake.
Green and colleagues (2002), however, found a more complicated pattern. Gender interacted
with other variables in predicting treatment initiation. Initiation was predicted in women by
alcohol diagnoses, while in men it was predicted by being employed or married. Failure to
initiate treatment was predicted in women by mental health diagnoses, but in men, by less
education. Treatment completion was predicted in women by higher income, in men, by
older age. Failure to complete was predicted in women by more dependence diagnoses and
higher (more negative) ASI Employment scores; in men, by worse psychiatric status,
receiving Medicaid, and motivation for entering treatment. More time spent in treatment
was predicted, in women, by alcohol or opiate diagnoses and legal/agency referral; in men,
by fewer mental health diagnoses, higher education, domestic violence victim status, and
prior 12-step attendance.

260

Phillips and colleagues (1995) performed a case-mix adjustment using age, gender, race,
education, mental health, drug use history, drug and mental health treatment history,
employment, and arrest history. Client severity at intake was a significant predictor of
outcomes three months after intake.
While the variables described above may be useful as a starting point for exploration, the
ideal combination of predictor variables may differ depending on the outcome and
performance measures to be adjusted. These predictors can only be selected and adequately
tested after the performance and outcome measures have been selected and data is available.
Conclusions and Recommendations
This chapter has outlined a number of performance and outcome measures with various
advantages and disadvantages. Ideally several complementary measures would be used as a
package to offset the individual weaknesses of each measure. One possible combination is
treatment show rates, treatment initiation within 14 days, treatment engagement within 30
days, CalOMS outcome measures and pre-post arrests. All of these measures have potential
and if all measures were used, the package of measures would monitor performance at the
beginning of the process (treatment show rates, treatment initiation within 14 days), during
treatment (treatment engagement within 30 days, arrests), and outcomes at treatment
discharge (CalOMS discharge outcome variables), and after treatment (arrests).
Treatment show rate stands out as the single measure that would take the least work to
implement (indeed, ADP already uses a similar measure for OTP funding). All of the other
measures would require significant but not insurmountable further work before being
deployed. Logical next steps would include the collection of feedback on these measures
from stakeholders as well as ideas for alternative measures, final selection of measures,
preparation of a plan for developing and deploying these measures, and developing case mix
adjustments where necessary.
For many of the measures discussed it would be extremely helpful to collect identifying
information on Prop 36 participants in all 58 counties based on data from county sources. A
set of information on each offender (e.g. name, date of birth, sex, social security number) and
their date of Prop 36 entry, would allow linkage to administrative databases (e.g. data from
DOJ, CDCR, CalOMS, health, mental health, employment, welfare, etc.) to identify and
track outcomes among offenders who entered Prop 36. This would fill in difficult data
“blind spots” where underreporting or inherent dataset limitations (e.g. preclude the tracking
of all Prop 36 participants).
It would also be helpful to collect information on services delivered to clients.
Unfortunately, to date, the only information generally collected on treatment services is
service modality (e.g. residential, outpatient). Some databases, such as the National Survey
of Substance Abuse Treatment Services (N-SSATS) indicate at the provider level whether
certain services are available (e.g. family counseling, HIV testing, individual therapy, drug
testing, etc.), but such databases cannot be used to determine whether any particular client
actually received these services. This creates significant challenges in using existing
measures to improve treatment and promote accountability. At a minimum, daily treatment
documentation should include the type, number and duration of standard counseling services

261

received, including individual counseling, group education, and group therapy. In addition,
as described above, staff ratings of subject attention and participation could be obtained at
each therapeutic sessions during the day. Development of a within treatment-day profile of
subject activity also is useful for quantifying the non-routine services that subjects receive
only from time to time. Referrals to social services, vocational services, onsite or offsite
medical and psychiatric services, or other ancillary appointments often are recorded as “no
shows” on routine group treatment documentation, so a more comprehensive log of daily
subject activity would provide a more accurate picture of treatment processes.
Performance and outcome measures hold substantial promise for monitoring and improving
Prop 36 performance and outcomes. However, if used improperly or without addressing the
significant data limitations, incentive issues, and other disadvantages associated with each
measure, inaccurate data and unintended consequences may cause the effort to do more harm
than good. Caution and careful research is urged as measures are selected and deployed.
References
Bond, G.R., et al. (2000). Measurement of fidelity in psychiatric rehabilitation. Mental
Health Services Research, 2, 75-87.
Brecht, M.L., Greenwell, L., & Anglin, M.D. (2005). Methamphetamine treatment:
trends and predictors of retention and completion in a large state treatment system
(1992-2002). Journal of Substance Abuse Treatment, 29, 295-306.
Butzin, C.A., Saum, C.A., & Scarpitti, F.R. (2002). Factors associated with completion
of a drug treatment court diversion program. Substance use & misuse, 37, 1615-33.
Friedmann P.D., Taxman F.S., & Henderson C.E. (2007). Evidence-based treatment
practices for drug-involved adults in the criminal justice system. Journal of
Substance Abuse Treatment, 32, 267 -277.
Garnick, D.W., Hodgkin, D., & Horgan, C.M. (2002a). Selecting data sources for
substance abuse services research. Journal of Substance Abuse Treatment, 22, 11–
22.
Gerstein, D.R., & Harwood, H.J. (1990). Treating drug problems: A study of the
evolution, effectiveness, and financing of public and private drug treatment systems
(vol. 1). Washington, DC: Institute of Medicine. National Academy of Press.
Gerstein, D.R., & Lewin, L.S. (1990). Treating drug problems. New England Journal of
Medicine, 323, 844 -848.
Green, C.A., et al. (2002). Gender differences in predictors of initiation, retention, and
completion in an HMO-based substance abuse treatment program. Journal of
Substance Abuse Treatment, 23, 285-95.
Harris, A.H.S., McKellar, J.D., & Saweikis, M. (2005). VA Care for Substance Use
Disorder Clients: Indicators of Facility and VISN Performance (Fiscal Years 2003
and 2004). Palo Alto, CA: Program Evaluation and Resource Center and HSR&D
Center for Health Care Evaluation. Avalable at:
www.chce.research.med.va.gov/chce/pdfs/2003PIG.pdf.

262

Hiller, M.L., Knight, K., & Simpson, D.D. (1999). Risk Factors That Predict Dropout
From Corrections-Based Treatment for Drug Abuse. The Prison Journal, 79, 411430
Hillhouse M.P., Marinelli-Casey P., Gonzales R., Ang A., Rawson R.A. (2007).
Predicting in-treatment performance and post-treatment outcomes in
methamphetamine users. Methamphetamine Treatment Project Corporate Authors.
Addiction, 102, 84 -95.
Hillhouse, M.P., Marinelli-Casey, P., Gonzales, R., Ang, A., Rawson, R.A. (2000).
Predicting in-treatment performance and post-treatment outcomes in
methamphetamine users. Methamphetamine Treatment Project. Addiction, 102, 84 95.
Institute of Medicine. Crossing the Quality Chasm: A New Health System for the 21st
Century. Washington, DC: National Academies Press, 2001. Available at:
www.nap.edu/books/0309072808/html/1.html.
Koenig, L., Fields, E.L., Dall, T.M., Ameen, A.Z., & Harwood, H.J. (2000). Using CaseMix Adjustment Methods to Measure the Effectiveness of Substance Abuse
Treatment: Three Examples Using Client Employment Outcomes. Prepared under
the NEDS contract by the Lewin Group.
Maltz, M.D. (2001). Recidivism. Orlando, Fl: Academic Press, Inc. Retrieved October
28, 2007 from : http://www.uic.edu/depts/lib/forr/pdf/crimjust/recidivism.pdf
McCamant, L.E., et al. (2007). Prospective validation of substance abuse severity
measures from administrative data. Drug and Alcohol Dependence, 86, 37-45.
McCollister, KE, & French, MT. (2003). The relative contribution of outcome domains in
the total economic benefit of addiction interventions: a review of first findings.
Addiction, 98, 1647-1659.
McCorry, F., et al. (2000). Developing performance measures for alcohol and other drug
services in managed care plans. Joint Commission on Quality Improvement, 26, 633643.
McClellan, A.T., Chalk, M., & Bartlett, J. (2007). Outcomes, performance, and quality –
What’s the difference? Journal of Substance Abuse Treatment, 32, 331-340.
McClellan, A.T., et al. (2005). Reconsidering the evaluation of addiction treatment:
From retrospective follow-up to concurrent recovery monitoring. Addiction, 100,
447–458.
National Committee for Quality Assurance (NCQA). HEDIS–Health Plan Employer
Data and Information Set, Vol. 2: Technical Specifications. Washington, DC:
NCQA, 2006. Available at:
www.ncqa.org/communications/publications/hedispub.htm.
Phillips C.D., Hubbard R.L., Dunteman G., Fountain D.L., Czechowicz D., & Cooper
J.R. (1995). Measuring program performance in methadone treatment using intreatment outcomes: an illustration. Journal of Mental Health Administration. 22,
214-225.
263

Simpson D.D., & Joe G.W. (2004). A longitudinal evaluation of treatment engagement
and recovery stages. Journal of Substance Abuse Treatment, 27, 89-97.
Substance Abuse and Mental Health Services Administration (SAMHSA) (2005).
National Outcome Measures (NOMs). Available at:
www.nationaloutcomemeasures.samhsa.gov.
Substance Abuse and Mental Health Services Administration (SAMHSA) (2007).
National Outcome Measures (NOMs). 2007. Available at:
http://www.nationaloutcomemeasures.samhsa.gov/outcome/index_2007.asp

264

GLOSSARY
Glossary 1: Glossary of Terms
Addiction Severity Index (ASI) – A standardized assessment designed to gather data on
treatment client status in seven domains: drug use, alcohol use, employment, family and
social relationships, legal status, psychiatric status, and medical status.
Board of Prison Terms (BPT) – The agency that protects public safety through the exercise
of its statutory authorities and policies, while ensuring due process to all criminal
offenders who come under its jurisdiction. The Board is responsible for the adjudication
of parole violations referred by the Parole and Community Services Division of the
California Department of Corrections. This agency developed the initial procedure for
referring and monitoring parolees during Prop 36’s first year.
Drug Court – Courts that oversee drug-using offenders in an approach emphasizing
treatment and close supervision; direct contact between judge and offender; and
collaboration between judge, prosecutor, defense attorney, and treatment provider.
Median – The “middle case” in an ordered distribution
Multivariate regression – Prediction of a dependent variable (e.g. treatment completion) by
two or more independent variables (e.g. primary drug and years of use).
N - The number of observations (e.g., people) in a statistical sample. In other words, the
sample size.
Parole and Community Services Division (P&CSD) of the California Department of
Corrections – The agency providing field supervision of California parolees.
Standard Deviation (SD) – Standard deviation is a measure of the spread or dispersion of a
set of data. It is calculated by taking the square root of the variance.

265

Glossary 2: Glossary of Abbreviations
ADP – California Department of Alcohol and Drug Programs
ADPA – Alcohol and Drug Program Administration
AOD – Alcohol and Drug
ASAM–PPC – American Society of Addiction Medicine Client Placement Criteria
ASI – Addiction Severity Index
BRAG – Behavior Response and Adjustment Guide
BTC – Break the Cycle
CADDS – California Alcohol and Drug Data System
CalDATA – California Drug and Alcohol Treatment Assessment
CalOMS – California Outcomes Measurement System
CalTOP – California Treatment Outcome Project
CalWORKs – California Work Opportunity and Responsibility to Kids
CASCs – Community Assessment and Service Centers
CATES – California Addiction Training and Education Series
CDCR – California Department of Corrections and Rehabilitation
CJ – Criminal Justice
COMP – California Opioid Maintenance Providers
COMPAS – Correctional Offender Management Profiling for Alternative Sanctions
CSAT – Center for Substance Abuse Treatment
CQI – Continuous Quality Improvement
CSAT – Center for Substance Abuse Treatment
DATA – Drug Addiction Treatment Act
DEA – Drug Enforcement Administration
DHS – California Department of Health, which has since been reorganized into the
Department of Public Health and Department of Health Care Services
DID – Difference in Differences
DMH – California Department of Mental Health
DOJ – California Department of Justice
DSM–IV–TR – Diagnostic and Statistical Manual of Mental Disorders–IV–Text
Revision
DTAP – Drug Treatment Alternative–to–Prison program

266

FY – Fiscal Year
EDD – California Employment Development Department
GED – General Educational Development
HEDIS – Health Plan Employer Data and Information Set
HOPE Probation – Hawaii’s Opportunity Probation with Enforcement
IDDT– Integrated Dual Diagnosis Treatment
LA CADA – Los Angeles Centers for Alcohol and Drug Abuse
LAAM – Levo–Alpha–Acetylmethadol
LSI–R – Level of Service Inventory–Revised
MHSA – Mental Health Services Act
MHSOAC – Mental Health Services Oversight and Accountability Commission
NCQA – National Committee for Quality Assurance
NIATx – Network for the Improvement of Addiction Treatment
NIDA – National Institute on Drug Abuse
NIH – National Institutes of Health
NIMBY – Not in My Back Yard
NOM – National Outcome Measures
NOS – Not Otherwise Specified
NQF – National Quality Forum
NRT – Narcotic Replacement Therapy
N–SSATS – National Survey of Substance Abuse Treatment Services
NTP – Narcotic Treatment Programs
OTP – Offender Treatment Program
PDSA – Plan–Do–Study–Act
PSATTC – Pacific Southwest Addiction Technology Transfer Center
RANT – Risk and Needs Triage
RCT – randomized controlled trial
SACPA – Substance Abuse and Crime Prevention Act
SAMSHA – Substance Abuse and Mental Health Services Administration
SCADP – Southern California Alcohol and Drug Programs, Inc.
SMRS – Social Model Recovery Systems, Inc.
SRIS – SACPA Reporting Information System
SSI/SSP – Supplemental Security Income/State Supplemental Program
267

STAR – Treatment’s Strengthening Treatment Access and Retention
TOPPS II – Treatment Outcomes and Performance Pilot Study
TSI – Treatment System Impact and Outcomes of Prop 36
UCLA – University of California Integrated Substance Abuse Programs
USFDA – U.S. Food and Drug Administration
VA – Department of Veterans’ Affairs

268

APPENDICES
Appendix A: UCLA’s Proposition 36 Stakeholder Survey
Participants
Respondents (n=290) in all 58 counties were asked to complete the 2007 Prop 36
Stakeholder survey by email. By November 21, 2007 UCLA had received completed or
partially completed surveys from 54 counties, which represent 93% of California’s 58
counties. Response rates by agency varied, with the greatest number of responses coming
from lead agencies (48 counties responded), followed by probation (35), court administrators
(27), public defenders (23), and district attorneys (19). Individual item response rates were
lower in part because stakeholders lacked time, did not have the information readily
available, or handled so few Prop 36 clients during the time period covered by the survey
that prospective respondents felt many questions were not applicable or determined that the
resources required to complete the survey outweighed perceived benefits.
Surveys
Surveys were designed by UCLA to address evaluation research questions agreed upon with
ADP. Draft copies of the stakeholder survey were sent to representatives from each
stakeholder group for feedback, and UCLA revised the instrument where appropriate.
As noted above, UCLA divided the survey into five distinct sections corresponding to
agencies involved in Prop 36: the lead agency (most often the county alcohol and drug
administrator), court administration, district attorney, public defender, and probation.
Questions focused on Prop 36 operation and needs; perceived strengths and weaknesses of
Prop 36 in each county; needs and services available to special populations (e.g., mentally ill,
homeless, high risk offenders); offender management strategies and other responses by the
criminal justice and treatment systems; and suggestions for improving Prop 36 treatment,
supervision, and operation.
All surveys were formatted as Microsoft Word Forms, which participants could complete
and return electronically. Additional copies were made available on the internet. Upon
request, paper copies of the surveys were made available.
Procedures
The survey along with a cover letter was emailed to the designated primary Prop 36 contact
for each county on July 30, 2007. Follow-up phone calls were placed to ensure that the
survey was received and to answer any questions about it. The survey was re-emailed as
needed to individuals who reported not receiving the survey previously.
Respondents were sent a letter thanking them for their participation and, if allowed, a $25
money order.

269

UCLA’s 2007 PROPOSITION 36 SURVEY: COUNTY LEAD AGENCY SECTION
•
•

Please note: If you do not have records indicating the actual number for each question,
please provide your best estimate.
In this survey, the term “Assessment” refers to the initial screening of offenders in order
to determine appropriate services and placement for that client. It does not refer to
subsequent assessments occurring at the treatment provider level.
In your county
from 7/1/05 to
6/30/06…

In your county
from 7/1/06 to
6/30/07…

1. How many offenders became eligible for the Prop
36 program? (probationers and parolees)
2. How many offenders opted for (were referred to)
the Prop 36 program, whether they completed their
assessment or not, and whether they actually
entered treatment or not?
3. How many offenders completed their Prop 36
assessment and were referred to treatment (whether
they actually entered treatment or not)?
4. How many Prop 36 offenders completed the Prop
36 program (i.e., completed requirements of both
treatment and probation)?
5. While waiting to enter treatment, were Prop 36
In your county from
offenders required to:
7/1/06 to 6/30/07
Not applicable/ no wait
5a. Be drug tested?
No
Yes
Not applicable/ no wait
5b. Attend self-help groups?
No
Yes
Not applicable/ no wait
5c. Enter an alternate level of care?
No
Yes
Not applicable/ no wait
5d. Attend drug education?
No
Yes
Not applicable/ no wait
5e. Attend Prop 36 orientation?
No
Yes
5f. Do something else? If yes, please
Not applicable/ no wait
No
Yes
describe
6. Was the initial drug and alcohol assessment for Prop 36
offenders conducted after sentencing? (not preNo
Yes
sentencing)
7. Were Prop 36 assessments conducted at the courthouse
where the offender was sentenced, or within walking
No
Yes
distance?
7a. If no, what percentage of Prop 36 offenders was driven from
%
court to the screening/assessment location at county expense?
7b. If no, what percentage of Prop 36 offenders was given bus
passes, vouchers, or other means of transporting themselves to
%
the assessment site?
8. Were Prop 36 offenders typically escorted to
No
Yes
assessment by a county employee or designate?

270

Appointment
9. Were the initial Prop 36 initial assessments
scheduled by appointment, or were walk-ins
Walk-ins allowed
allowed? (please check both boxes if both strategies
were used).
10.
How many visits were normally required to
complete the Prop 36 assessment (see top of page 1
visits
for definition of assessment)?
11.
Was a formal standardized orientation session,
explaining processes and obligations associated
No
Yes
with the Prop 36 program, routinely provided to
most offenders?
11a. If yes, what entity or entities provided the orientation? (judge, probation
officer, treatment provider etc.)
If yes, at what point(s) was this orientation
provided?
11b. Prior to sentencing
No
Yes
11c. During the sentencing hearing
No
Yes
11d. During the initial assessment process
No
Yes
11e. After assessment but prior to treatment
No
Yes
admission
11f. During or after treatment admission
No
Yes
12.
Were the following practices used to address Prop 36 offender motivation level?
12a. Motivational interviewing
No
Yes
12b. Denial management sessions
No
Yes
12c. A peer-based buddy system
No
Yes
12d. Pre-treatment education sessions
No
Yes
12e. Other. If yes, describe
No
Yes
13.
Was a mental health screening routinely conducted
No
as part of the normal Prop 36 assessment (Do not
include the ASI or the ASAM-PPC)?
13a. If No, was a mental health screening conducted in
response to certain outcomes in the initial
No
assessment (for example, triggered by answers to
certain items on the ASI or ASAM-PPC)?
13b. If yes to 13 or 13a, what instrument was used?
If yes to 13 or 13a, were offenders with a mental disorder usually:
13c. Assigned to a treatment program that
No
specialized in treating co-occurring drug
abuse and psychiatric disorders?
13d. Given a referral for mental health services at
a provider other than the program they are
No
being referred to for substance abuse
treatment.
13e. Prescribed psychiatric medication?
No
271

Yes

Yes

Yes

Yes
Yes

13f. Seen by a licensed mental health
professional?
13g. Other strategies? If yes, please describe

No

Yes

No

Yes

No

Yes

No

Yes

14b. Provided with housing assistance or
placement?

No

Yes

14c. Referred to treatment program specializing in
homeless clients?

No

Yes

No

Yes

No

Yes

No

Yes

No

Yes

No

Yes

14.
Were special strategies in place for homeless
Prop 36 offenders? If yes were homeless offenders
usually:
14a. Referred to residential treatment?

14d. Other strategies? If yes, please describe
15.
Were special strategies in place for Prop 36
offenders with many prior convictions? If yes, were
these offenders usually:
15a. Subject to increased monitoring / supervision.
15b. Placed in residential treatment?
15c. Other strategies? If yes, please describe

16.
Approximately what percentage Prop 36 offenders who completed
%
treatment participated in a continuing care (a.k.a. aftercare) program?
17.
Did Prop 36 offenders receive services relating
No
Yes
to employment, vocational skills, or job training?
If yes, could services usually received be described as:
17a. Job readiness assessment?
No
Yes
17b. GED education?
No
Yes
17c. Vocational counseling?
No
Yes
17d. Job-seeking skills training?
No
Yes
17e. Resume assistance?
No
Yes
17f. Job skills training?
No
Yes
17g. Job leads (information on job
No
Yes
openings)?
17h. Job placement?
No
Yes
17i. Other? If yes, please describe
No
Yes
If employment/vocational services were provided to Prop 36 clients:
17j. Were they typically provided by a drug
Drug treatment
treatment program or by a separate
Specialized
program specializing in employment /
employment / vocational
vocational services (and not a drug
program
treatment program)
272

17k. On-site at the offender’s primary drug
On-site
treatment location or off-site at a different
Off-site
location?
Did the Prop 36 lead agency or Prop 36 alcohol and other drug office participate
18.
in the following countywide planning processes:
18a. Proposition 63 implementation planning?
No
Yes
18b. AB 636 Child Welfare Services program
No
Yes
improvement?
18c. Proposition 10 “0” to “5” planning?
No
Yes
18d. Local Workforce Investment Board
No
Yes
planning?
Was there a countywide policy guiding Prop 36 drug testing practices by:
19.
19a. Treatment?
No
Yes
If yes, did the policy specify that drug tests be conducted:
19a1. At random?
No
Yes
19a2. On a regular schedule?
No
Yes
19a3. On suspicion of use?
No
Yes
19a4. For other reasons? If yes please describe:
No
Yes
19b. Probation?
If yes, did the policy specify that drug tests be conducted:
19b1. At random?
19b2. On a regular schedule?
19b3. On suspicion of use?
19b4. For other reasons? If yes please describe:

No

Yes

No
No
No

Yes
Yes
Yes

No

Yes

positive tests
How many positive drug tests were typically
20.
allowed before a Prop 36 offender was subject to
Varies widely by case.
consequences?
21.
Did all programs of the same modality use an
agreed-upon definition of “treatment completion”
No
Yes
that is more detailed than the CalOMS definition
(“completed treatment/recovery plan and/or goals”)?
22.
Were performance based contracts with Prop 36
No
Yes
treatment providers (i.e., level of payment is based on
performance measures) used?
23.
Did referred, assessed, and placement counts reported by your county to the
SACPA Reporting Information System (SRIS) represent offenders (people counted
once even if referred, assessed, or placed more than once) or events (each referral,
assessment, or placement is counted)?
Offender
Events
Other
s
23a. Referred: If “other” please describe:
23b. Assessed: If “other” please describe:

273

23c. Placed: If “other” please describe:
Did counts of referrals, assessments, and
24.
placements reported to SRIS include parolees sent
to Prop 36 by a parole agent?

Yes for Referrals
Yes for Assessments
Yes for Placements
No, not for any of the
above

25.
Did your county receive Offender Treatment
No
Yes
Program funds?
25a. If yes, were all activities proposed in your
All
OTP application fully implemented as of
Some being
activities
6/30/2007, or were some still being
developed
implemented
developed?
26.
Do you have any additional comments regarding Prop 36, or ideas for improving its
implementation?
27.
Were any questions on this survey unclear or difficult to answer? Are there
questions that you think should be asked in the future that were not covered in this
survey?
28.
Please provide your contact information below.
Name:
Job Title:
County Name:
Address:
Phone:
Email:
Fax:
29.

Would you like to receive a $25 money order for completing this survey?
No
Yes
29a. If money order should be addressed to someone different than the person listed
above, please specify (otherwise leave blank):
Name:
Job Title:
County Name:
Address:
Phone:
Email:
Fax:

After making a copy of this survey for your records, please send this survey to Liz Evans
by email to laevans@ucla.edu, by fax to (310) 473-7885, or by mail to UCLA, 1640 S.
Sepulveda Blvd. Ste. 200, Los Angeles, CA 90025.
Thank you!

274

UCLA’s 2007 PROPOSITION 36 SURVEY: PROBATION SECTION
These questions ask about the status of Prop 36 offenders in your county. If you do not
have records indicating the actual number for each question, please provide your best
estimate.
Between July 1, 2006 and June 30, 2007 in your county…
1. How many Prop 36 offenders were on probation? Please
include offenders placed on probation during the year and
offenders
those already in Prop 36 before July 1, 2006.
1a. How many of these were on formal probation?
offenders
1b. How many of the total in question 1 acquired at least
offenders
one new drug violation while in Prop 36?
1c. How many of the total in question 1 were revoked from
Prop 36 probation and re-sentenced? (not placed back
offenders
into Prop 36)
1c1. How many were sent to jail or prison as a result of
offenders
revocation?
2. How many offenders completed their Prop 36 probation
offenders
term?
3. Please describe general Prop 36 reporting by treatment providers to probation:
Never

Almost
Never

Some
times

Almost
Always

Always

Never

Almost
Never

Some
times

Almost
Always

Always

3a. Treatment plans were reported by
treatment providers within 30 days of
notice from probation that the provider
has been designated to provide drug
treatment
3b. Positive/missed drug tests were
reported by treatment providers within
2 weeks after test date
3c. Other noncompliance was reported by
treatment providers within 2 weeks
after noncompliance occurred
3d. Quarterly progress reports were sent by
treatment providers within 2 weeks
after the end of the quarter
3e. Successful treatment completion was
reported by treatment providers within
2 weeks
3f. Treatment drop-out was reported by
treatment providers within 2 weeks
4. Were the results of positive probation drug
test results shared with the offender’s
treatment provider(s)?
275

5. Among offenders who opted for Prop 36 but did not enter treatment, what proportion
would you estimate did not do so for the following reasons? (if offenders did not enter
treatment for more than one reason, percentages may add to more than 100%)
5a. Offender was re-arrested shortly after sentencing.

%

5b. Offender changed mind about participating after learning more
about the Prop 36 requirements.

%

5c. Offender never intended to enter treatment.

%

5d. Offender started using drugs again.

%

5e. Offender couldn’t afford fees required to enter treatment.

%

5f. Prop 36 requirements were incompatible with other obligations
(work schedule, for example).

%

5g. Other (describe

%

)

6. Did you have Prop 36 dedicated probation officers
(who specialized in or handled Prop 36 cases only)?

No

6a. If yes, how many Prop 36 cases were handled by
Prop 36 dedicated probation officers?
6b. If yes, what was the typical caseload (only Prop
36 cases) for Prop 36 dedicated probation
officers?
6c. If no, what was the typical caseload (both Prop 36
and non-Prop 36 cases) for probation officers who
handled Prop 36 cases?
7. What was the total amount originally assessed in
probation fees/fines per Prop 36 offender, on average?
7a. What percentage of Prop 36 offenders paid at
least some of their probation fees/fines?

Yes

cases
cases

cases

$
%

8. What were the consequences for Prop 36 offenders who met all
Prop 36 program requirements, but were unable to pay their
probation fees/fines?
8a. Part of the fee was waived / reduced

No

Yes

8b. All of the fee was waived

No

Yes

8c. Payment plan was set up

No

Yes

8d. Offender was assigned to community
service

No

Yes

8e. Offender remained on probation

No

Yes

8f. Other. If yes, please describe:

No

Yes

276

9. Do you have any other comments on Prop 36 or ideas for improving its
implementation?

10. Were any questions on this survey difficult to answer? Are there questions that you
think should be asked in the future that were not covered in this survey?

11. Please provide your contact information below.
Name:
Job Title:
County Name:
Address:
Phone:
Email:
Fax:
12. Would you like to receive a $25 money order for completing this survey?
Yes

No

12a. If money order should be addressed to someone different than the person listed
above, please specify (otherwise leave blank):
Name:
Job Title:
County Name:
Address:
Phone:
Email:
Fax:

After making a copy of this survey for your records, please send this survey to Liz Evans
by email to laevans@ucla.edu, by fax to (310) 473-7885, or by mail to UCLA, 1640 S.
Sepulveda Blvd. Ste. 200, Los Angeles, CA 90025.
Thank you.

277

UCLA’s 2007 PROP 36 SURVEY: COURT ADMINISTRATION SECTION
Between July 1, 2006 and June 30, 2007 in your county…
1. Please indicate whether the following procedures were used to handle Prop 36 cases:
1a. Dedicated/centralized court for all Prop 36
No
Yes
offenders.
1b. Dedicated/centralized court for some Prop 36
No
Yes
offenders.
1c. Drug court setting (same bench officer sees both
Prop 36 and Drug Court cases) for all Prop 36
No
Yes
offenders.
1d. Drug court setting (same bench officer sees both
No
Yes
Prop 36 and Drug Court cases) for some Prop 36
offenders.
1e. Expedited case processing.
No
Yes
1f. Case conferences.
No
Yes
1g. Probation assessment hearings.
No
Yes
1h. Status hearings.
No
Yes
1i. Tailored drug testing requirements.
No
Yes
1j. Other. If yes, please describe:
No
Yes
2. Were procedures for managing Prop 36 offenders the
Procedures were the
same across courts or did procedures vary
same across courts
substantially?
Procedures varied
across courts
3. Did the court assign Prop 36 offenders to receive these services if they were needed:
3a. Employment services?
No
Yes
3b. Literacy training?
No
Yes
3c. Mental health services?
No
Yes
3d. Family reunification or family dynamics
No
Yes
counseling?
3e. Housing for homeless offenders?
Yes
No
3f. Other services besides drug treatment? If yes,
No
Yes
please describe:
4. Did the court ever assign Prop 36 opiate
users to Narcotic Replacement Therapy?
No
Yes
(NRT – Methadone maintenance, for
example. Not detox-only)
4a. If yes, what criteria did the court use to
Only if drug free treatment
decide whether to use narcotic
was unsuccessful
replacement therapy?
Only if drug free treatment
was unavailable
As the first option for treating
opiate users
Other, please describe:

278

4b. If no, what were the reason(s)?

5.

6.

7.

8.

Not offered to Prop 36 offenders by
county policy
Narcotic Replacement Therapy was
unavailable
Philosophical opposition to
narcotic replacement therapy
Other, please describe:
Were the following strategies used by the court in response to offender noncompliance
with the Prop 36 program?
5a. Increased level of supervision
No
Yes
5b. Increased frequency of drug testing
No
Yes
5c. Changed treatment level of care or treatment
No
Yes
length
5d. Community service
No
Yes
5e. Offender required to observe Prop 36 courtroom
No
Yes
proceedings
5f. Fines
No
Yes
5g. Bench warrants
No
Yes
5h. Writing assignments
No
Yes
5i. Other. If yes, please describe:
No
Yes
Were the following strategies used by the court in response to offender compliance with
the Prop 36 program?
6a. Decreased level of supervision
No
Yes
6b. Decreased frequency of drug testing
No
Yes
6c. Changed treatment level of care or treatment
No
Yes
length
6d. Graduation ceremonies
No
Yes
6e. Certificates of completion
No
Yes
6f. Gift certificates or vouchers
No
Yes
6g.Verbal praise or recognition
No
Yes
6h. Candy
No
Yes
6i. Other? If yes, please describe:
No
Yes
What was the total amount assessed in court
$
fees/fines per Prop 36 offender, on average?
%
7a. Approximately what percentage of Prop 36
offenders paid at least some of their court
fees/fines?
What were the consequences for Prop 36 offenders who met all Prop 36 program
requirements, but were unable to pay their court fees/fines?
8a. Part of the fee/fine was waived
No
Yes
8b. All of the fee/fine was waived
No
Yes
8c. Payment plan was set up
No
Yes
8d. Offender was assigned to community service
No
Yes
8e. Offender remained on probation
No
Yes
8f. Other. If yes, please describe:
No
Yes

279

9. How many months was the average Prop 36 offender
required to remain on probation after successfully
completing treatment?
10. Approximately what percentage of Prop 36 offenders
who successfully completed Prop 36 petitioned the
court for expungement/dismissal of their original
Prop 36 charge(s)?
11. Were any new offenses added to the list of offenses
that made an offender eligible for Prop 36 in 20062007?
11a. If yes, please list those offenses here.
12. On records sent to the Department of Justice, what
disposition codes or sentencing codes were used to
indicate an offender’s decision to participate in Prop 36?
13. Do you have any additional comments regarding Prop 36, or
implementation?

months
%

No

Yes

ideas for improving its

14. Were any questions on this survey unclear or difficult to answer? Are there questions that
you think should be asked in the future that were not covered in this survey?
15. Please provide your contact information below.
Name:
Job Title:
County Name:
Address:
Phone:
Email:
Fax:
16. Would you like to receive a $25 money order for completing this survey?
No
Yes
16a. If money order should be addressed to someone different than the person listed
above, please specify (otherwise leave blank):
Name:
Job Title:
County Name:
Address:
Phone:
Email:
Fax:
After making a copy of this survey for your records, please send this survey to Liz Evans
by email to laevans@ucla.edu, by fax to (310) 473-7885, or by mail to UCLA, 1640 S.
Sepulveda Blvd. Ste. 200, Los Angeles, CA 90025.
Thank you!
280

UCLA’s 2007 PROPOSITION 36 SURVEY: PUBLIC DEFENDER SECTION
Between July 1, 2006 and June 30, 2007 in your county…
1. Approximately what proportion of eligible offenders
declined the Prop 36 program?

%
1.

2. What were the top three reasons for declining the Prop
36 program?

2.
3.

3. Were reasons for declining the Prop 36 program
different from those listed above for offenders who
were:
3a.Homeless?

No

Yes

No

Yes

No

Yes

No

Yes

No

Yes

3a1. If yes, please describe:
3b. Mentally ill?
3b1. If yes, please describe:
3c. Both Homeless and Mentally ill?
3c1. If yes, please describe:
3d. Opiate users?
3d1. If yes, please describe:
4. Were Prop 36-eligible offenders commonly advised to
decline the Prop 36 program in favor of other options
(e.g., deferred entry of judgment, drug court,
incarceration)?
5. Approximately what percentage of Prop 36 offenders
were represented by a public defender or courtappointed attorney?
5a. What was the typical caseload for public defenders
or court-appointed attorneys who represented Prop 36
offenders?
6. Were there public defenders or court-appointed
attorneys who specialized in Prop 36?
6a. If yes, approximately what percentage of Prop 36
offenders were assigned to a public defender who
specialized in Prop 36?

281

%
Prop 36 cases
Non-Prop 36 cases
No

%

Yes

7. Do you have any additional comments regarding Prop 36, or ideas for improving its
implementation?

8. Were any questions on this survey unclear or difficult to answer? Are there questions
that you think should be asked in the future that were not covered in this survey?
9. Please provide your contact information below.
Name:
Job Title:
County Name:
Address:
Phone:
Email:
Fax:
10. Would you like to receive a $25 money order for completing this survey?
Yes

No

10a. If money order should be addressed to someone different than the person listed
above, please specify (otherwise leave blank):
Name:
Job Title:
County Name:
Address:
Phone:
Email:
Fax:

After making a copy of this survey for your records, please send this survey to Liz Evans
by email to laevans@ucla.edu, by fax to (310) 473-7885, or by mail to UCLA, 1640 S.
Sepulveda Blvd. Ste. 200, Los Angeles, CA 90025.
Thank you!

282

UCLA’s 2007 PROPOSITION 36 SURVEY: DISTRICT ATTORNEY SECTION
Between July 1, 2006 and June 30, 2007 in your county…
1. What Prop 36-specific policies were in effect:
1a. Standard set of charges on which offenders were eligible
for Prop 36?
1b. Charging practices designed for Prop 36?
If yes, please describe:
1c. Case processing designed for Prop 36?
If yes, describe:
1d. Plea negotiation guidelines designed for Prop 36?
If yes, describe:
1e. Plea agreements under which Prop 36-eligible
defendants could decline Prop 36?
1f. Other?
If yes, describe:
2. What was the typical caseload for prosecutors who worked
on Prop 36 cases?
3. Did some prosecutors specialize in Prop 36 cases?
3a. If yes, what percentage of Prop 36 cases were assigned to
prosecutors who specialize in Prop 36?
3b. Did some prosecutors work exclusively on Prop 36
cases?

No

Yes

No

Yes

No

Yes

No

Yes

No

Yes

No

Yes

Prop 36
cases

Non-Prop
36 cases

No

Yes

% of cases
No

Yes

4. Do you have any additional comments regarding Prop 36, or ideas for improving its
implementation?

5. Were any questions on this survey unclear or difficult to answer? Are there questions that
you think should be asked in the future that were not covered in this survey?

283

6. Please provide your contact information below.
Name:
Job Title:
County Name:
Address:
Phone:
Email:
Fax:
7. Would you like to receive a $25 money order for completing this survey?
Yes

No

7a. If money order should be addressed to someone different than the person listed
above, please specify (otherwise leave blank):
Name:
Job Title:
County Name:
Address:
Phone:
Email:
Fax:

After making a copy of this survey for your records, please send this survey to Liz Evans
by email to laevans@ucla.edu, by fax to (310) 473-7885, or by mail to UCLA, 1640 S.
Sepulveda Blvd. Ste. 200, Los Angeles, CA 90025.
Thank you!

284

Appendix B: UCLA’s Proposition 36 Treatment Program Survey
Participants
UCLA selected a random sample of 150 Prop 36 providers who served more than 5 clients in
2005-2006 according to records in CADDS. Follow-up phone calls to all providers
determined that six were not (or were no longer) Prop 36 providers. Six additional treatment
providers were randomly selected as replacements. After calling these programs to verify
that they were serving Prop 36 clients, researchers were informed that one of the replacement
programs also did not serve Prop 36 clients. Continuing phone calls determined that three
additional providers did not serve Prop 36 clients, but due to time constraints, these were not
replaced. Therefore, surveys were sent to 146 Prop 36 treatment programs, of these, 86
responded (58.9%).
In addition, due to interest in maintenance treatment, UCLA oversampled NTP providers by
selecting 10 additional NTP providers that met selection criteria in 2005-2006. This was in
addition to the three providers that were included in the random sample described above.
Phone calls confirming that these NTP providers served Prop 36 clients determined that one
of these providers did not serve Prop 36 clients at the time of the survey. Therefore, the
oversampling effort resulted in 9 additional surveys sent to NTPs that served Prop 36
participants, of these, 5 responded (55.5%).
Survey Construction
Surveys were designed by UCLA to address evaluation research questions agreed upon with
the Department of Alcohol and Drug Programs. Draft copies of the treatment provider
survey were sent to two current and former treatment providers for feedback, and UCLA
revised the instrument where appropriate in response.
Questions focused on program characteristics; treatment services; treatment population; drug
testing, treatment capacity; treatment characteristics; and treatment completion. Respondents
were also asked if they had any additional comments/information regarding the
implementation/operation of Prop 36 in their treatment programs.
Procedures
Initial scripted phone calls were made from July to early August 2007 to confirm whether
programs served Prop 36 clients, verify mailing addresses, and obtain the names of the
program directors to whom the surveys were to be addressed. The surveys, along with cover
letter and payment form, were express mailed via DHL from July 30, 2007 to August 16,
2007. Programs with only P.O. Box delivery available were sent the surveys via First Class
U.S. Mail.
Follow-up calls were placed to ensure that the survey was received and to answer any
questions about it. The survey was re-mailed, faxed, or e-mailed as needed to individuals
who reported not receiving the survey previously. Additional calls were placed and e-mails
sent to non-respondents.
Along with the treatment program survey and cover letter, each program director was sent a
payment form that asked the participant if s/he wished to receive a $75 money order. If s/he
285

checked “yes,” s/he was asked to indicate to whom the money order should be made payable
and the mailing address. Out of the 91 programs with completed surveys, 69 accepted
payment, seven declined, and 15 did not include the form. Follow-up calls were made and
voice messages were left with those that did not include the payment form, asking them if
they wished to receive payment. Four of these programs confirmed that they wished to
receive payment and two of them declined payment. Therefore, a total of 73 programs
accepted and nine declined payment. The remaining nine that did not include the payment
form did not return phone messages left by study staff, or were not able to be reached.
Respondents were sent a letter thanking them for their participation and, if allowed, a $75
money order.
Study staff created the data entry database using Filemaker Pro v.5 database software. The
data entry fields were pre-tested and minor fixes were made. A notes section was added to
indicate any inconsistencies with the data (e.g., marking two responses, when only one
should have been marked; writing in a range [e.g., 3-5] instead of the average [e.g., 4]). Data
was entered, cleaned, and exported as a .csv file, and then formatted for SPSS.

286

PROPOSITION 36 EVALUATION

2007 Treatment Provider Survey
University of California, Los Angeles
Integrated Substance Abuse Programs (ISAP)
July 2007
This study is being conducted by the UCLA Integrated Substance Abuse Programs under a
contract with the California Department of Alcohol and Drug Programs. This Treatment
Provider Survey is an important component of UCLA’s statewide evaluation of Proposition
36.

Your participation is voluntary and confidential. All results will be reported only in
aggregated form (across programs) so that individual programs cannot be identified.
Upon receipt of your completed survey, we will send you a money order for $75 if your
program allows payment.
Please answer questions in this survey based on activities at:
CADDS / CalOMS Provider ID: [insert id]
Please answer based on activities occurring over the past fiscal year:
July 1, 2006 through June 30, 2007.
If the Provider ID above is not associated with your facility or if you have any questions,
please do not hesitate to contact Joy Yang at:
Tel: (310) 267-5252
Email: joyinla@ucla.edu
UCLA Integrated Substance Abuse Programs
1640 S. Sepulveda Blvd. Suite 200
Los Angeles, CA 90025
To participate and receive payment, please return your completed survey in the enclosed
pre-addressed postage paid DHL envelope by August 31, 2007.

287

Date survey was filled out:

/
/
month / day / year

Job title of person completing survey:
Program Director ......................................................1
Program Manager .....................................................2
Executive Director ....................................................3
Clinic Administrator .................................................4
Program Supervisor ..................................................5
Other .........................................................................6
Specify______________________________

Has this program provided treatment services to any Prop 36 clients in the past year?
______No

Stop! Disregard the rest of this survey and return
this form in the enclosed postage-paid envelope.

______Yes

Continue with the survey!

288

SECTION 1: PROGRAM CHARACTERISTICS
1.

How long has this program provided treatment services to Prop 36
clients?

2.

__________years

How many Prop 36 clients were admitted to your program in the past year?
(Please provide an estimate if you do not know the exact number.)
Number of Prop 36 clients

3.

What are this program’s days and hours of
operation (e.g., M-F, 9am-2pm; Sat., 9am-1pm)?

___________________

__________________________

4-7. Does this program offer these services during the day, in the evenings and/or on the
weekends?
a. During the day
No
Yes

b. In the evenings
No
Yes

c. On weekends
No
Yes

4. Individual sessions ______

______

______

______

______

______

5. Group sessions

______

______

______

______

______

______

6. Other services

______

______

______

______

______

______

7. Please describe______________________________________
8.

If your program provides outpatient treatment services, how many hours
per week are clients expected to attend treatment?
_______hours per week

9-15. Please fill in the number of program staff in each category providing services to
Prop 36 clients.
Number of staff

9.

Counselors (include related positions, e.g., therapists, case
managers, recovery specialists)

10.

Nurses

11.

Physicians

12.

Psychiatrists

13.

Psychologists

___________

14.

Social Workers

___________

15.

Other (e.g., Data entry staff, Administrative support, Aides)

___________
___________
___________

___________

16. What proportion of your program’s direct care staff hold certification in
substance abuse counseling (e.g., CAC, CADC or CADAC)?
____________%

289

17.

What proportion of your program’s direct care staff hold a master’s
degree or higher?
____________%

18. Approximately what percent of your staff are in recovery?
19.

To what extent is your program kept informed of the criminal justice status (e.g.,
revocation of probation, dismissal of case) of Prop 36 clients?
Not At
All
0

20.

____________%

Limited
Extent
1

Moderate
Extent
2

Great
Extent
3

Has your program been able to secure supplemental funding (other than SACPA
trust funds, SATTA, county general funds, and fees collected from Prop 36
clients) to facilitate the operation of Prop 36 (e.g., grants from private or federal
agencies, or the Offender Treatment Program)? _____No _____Yes

SECTION 2: TREATMENT SERVICES
1-34. What types of services have been available to Prop 36 clients in the past fiscal year
(7/1/06 to 6/30/07)? Please indicate if your program:
1. Provides this service on-site.
2. Refers clients to this service through a cooperative/formal agreement with
other service providers.
3. Does not provide this service on-site and does not have a formal referral
agreement with other programs.
1. Provided
on-site

2. Referred,
cooperative
agreement

3. Service
not
provided
/No formal
referral

Specific treatments & ancillary services:

1.

Family counseling

1

2

3

2.

Domestic violence counseling

1

2

3

3.

Parenting assistance

1

2

3

4.

Childcare

1

2

3

5.

Literacy training

1

2

3

6.

GED education

1

2

3

7.

Transportation assistance

1

2

3

8.

Drug/alcohol education

1

2

3

9.

Physical health (nurse or physician on-site)

1

2

3

10. HIV or Hepatitis C testing or prevention

1

2

3

11. Transitional housing

1

2

3

12. Sober living environment

1

2

3

290

1. Provided
on-site

2. Referred,
cooperative
agreement

3. Service
not
provided
/No formal
referral

13. Job readiness assessment

1

2

3

14. Vocational counseling

1

2

3

15. Job-seeking skills training

1

2

3

16. Résumé preparation assistance

1

2

3

17. Job skills training

1

2

3

18. Job leads

1

2

3

19. Job placement

1

2

3

1

2

3

1

2

3

1

2

3

1

2

3

1

2

3

25. Crisis intervention services
Behavioral interventions for mental health
26. problems (e.g., social skills training, symptom
management)
Outreach (services provided in the
27.
community)
Aftercare

1

2

3

1

2

3

1

2

3

28. In-person continuing care

1

2

3

29. Telephone-based continuing care

1

2

3

30. Follow-up counseling

1

2

3

31. Support groups

1

2

3

32. Social activities (e.g., alumni meetings)

1

2

3

33. Other

1

2

3

Vocational training/employment

Mental Health/Dual Diagnosis

20. Mental health assessment or diagnosis
Mental health counseling/therapy (group or
21.
individual)
Mental health medication services
22.
(e.g., prescription, monitoring)
Special "dual diagnosis" groups
23.
(includes dual diagnosis 12-step groups)
24. Psychiatric case management

34. Specify

________________________

291

35-37. Of the services listed above, what are the top three most urgent or pronounced
service needs for Prop 36 clients in your program? Enter the numbers to the left
of the service next to your 1st, 2nd and 3rd choice (e.g., enter 3 for Parenting
assistance).
35.
1st
_________
36.
2nd
_________
37.
3rd
_________
38. What percentage of Prop 36 clients received services related to employment,
vocational skills, or job training in the past year?
____________%
SECTION 3. TREATMENT POPULATION
1-6. Please provide information on your program’s Prop 36 treatment population in the
past year. If you do not know the exact percent please estimate the percentage in
each category or check the “unable to estimate” box.
Unable to
Percentage estimate
1.
What percentage of your program’s Prop 36 clients had a
co-occurring mental disorder at treatment entry
(i.e., schizophrenia, bipolar disorder, or severe anxiety
_________
disorders, in addition to a substance use disorder)?
%
□
2.
What percentage of your program’s Prop 36 clients were
_________
homeless at treatment entry (e.g., no permanent residence,
%
□
temporary placement, living on the street or in a car)?
3.
What percentage of your program’s Prop 36 clients were
homeless AND had a co-occurring mental disorder at
_________
treatment entry?
%
□
4.
What percentage of your program’s Prop 36 clients were
_________
receiving psychiatric medications at treatment entry
(e.g., antidepressants, antipsychotics, mood stabilizers)?
%
□
5.
6.

What percentage of your Prop 36 clients were opiate users
at treatment entry (e.g., heroin, oxycodone, morphine)?

_________
%

□

Of the Prop 36 clients who were opiate users, what
percentage receive a narcotic replacement medication
(e.g., Methadone, Buprenorphine, Subutex, Suboxone)?

_________
%

□

7. Is your program a Dual Diagnosis Treatment Program?
_____ No (skip to Question 12) _____Yes
If yes to Question 7, does your program:
8. Use an Integrated Dual Disorder Treatment Approach?
9. Conduct Psychodiagnostic Assessment (e.g., the SCID)?
DO NOT COUNT the ASI or the ASAM-PPC
10. Receive Prop 63 funding (Mental Health Services
11. Report client or outcome data to a State or County Mental
Health Database (e.g., Department of Mental Health)?

292

No

Yes

______

_______

______
______

_______
_______

______

_______

12.

Does your program refer Prop 36 clients with a co-occurring mental illness
to a Mental Health Treatment Provider?
_____No _____Yes

13.

Does your program treat Prop 36 clients who report
being homeless at treatment entry?
_____No (skip to Section 4) _____Yes
If Yes to Question 13, does your program:
14. Place them in residential treatment
15. Provide housing assistance
16. Attempt to find them supportive housing through AB
2034
17. Attempt to find them other stable housing
18. Provide other services not listed

No

Yes

_______
_______

_______
_______

_______
_______
_______

_______
_______
_______

19. Please describe: ___________________________________________
SECTION 4: DRUG TESTING
1. Does your program drug test Prop 36 clients?
____No (Skip to Question 22 on page 8) _____Yes
Does your program conduct:
2.

Random drug testing

_______No ______Yes

3.

Tests for cause (i.e., suspicion of drug use)

4.

Regularly scheduled drug tests (e.g., weekly, monthly) _______No ______Yes

_______No ______Yes

How does your program conduct drug testing? (check all that apply):
5. ___ Sample is collected at the treatment facility, instant results.
6. ___ Sample is collected at the treatment facility, sent to lab for analysis (no
instant results).
7. ___ Sample is collected at and analyzed at a site outside of the treatment facility.
8-10. On average, how often are Prop 36 clients typically tested for drug use by this
treatment program per month? Please fill in your answers in the column(s) that
correspond to the modality(s) this program provides. For example, if this program
provides both Outpatient and Residential Treatment, please complete the first two
columns and leave the last column blank.
10. Narcotic Treatment
8. Outpatient Treatment
9. Residential Treatment
Program

_____________________
Number of drug tests

_______________________
Number of drug tests

293

_______________________
Number of drug tests

11.

Does your program report all positive drug tests to criminal justice personnel (e.g.,
courts, probation) for Prop 36 clients who are on probation?
_______No ______Yes (skip to Question 15)

12-14. How many positive drug tests typically occur for Prop 36 clients on probation
before test results from this program are reported to criminal justice personnel?
12. Outpatient
Treatment

13. Residential
Treatment

____________________
Number of positive tests

_____________________ ______________________
Number of positive tests
Number of positive tests

14. Narcotic Treatment Program

15. Does your program report all positive drug tests to criminal justice personnel (e.g.,
courts, parole) for Prop 36 clients who are on parole?
_______No ______Yes (skip to Question 19)
16-18. How many positive drug tests typically occur for Prop 36 clients on parole before
test results from this program are reported to criminal justice personnel?
16. Outpatient
Treatment

17. Residential Treatment

18. Narcotic Treatment
Program

____________________
Number of positive tests

_____________________
Number of positive tests

______________________
Number of positive tests

19-21. How many positive drug tests typically occur before a Prop 36 client is
discharged from this program?
19. Outpatient
Treatment

20. Residential
Treatment

21. Narcotic Treatment Program

____________________
Number of positive tests

____________________
Number of positive tests

__________________________
Number of positive tests

22.

Does your program receive Prop 36 clients’ drug test results from criminal
justice personnel (e.g., probation, parole)?
______No _____Yes

294

23-29. To what extent are the following consequences given to Prop 36 clients who test
positive for drugs at your program?
Not
Limited Moderate Great
At All Extent
Extent
Extent
23. An adjustment is made to the client’s
treatment plan (e.g., participation in
groups and/or 12-step meetings is
increased).
24. A change is made in the client’s level of
care (e.g., transferred from outpatient to
intensive day treatment).
25. The frequency of drug testing is
increased.
26. The client is discharged with a referral to
another program.
27. The client is discharged without a referral
to another program.
28. Other consequences

0

1

2

3

0

1

2

3

0

1

2

3

0

1

2

3

0

1

2

3

0

1

2

3

29. Describe_____________________________________________
30. To what extent are rewards/positive incentives given to Prop 36 clients testing
negative for drugs at your program?

Not At
All
0

Limited
Extent
1

Moderate
Extent
2

Great
Extent
3

SECTION 5: TREATMENT CAPACITY
1-3. What has been the average number of days from Prop 36 assessment to entering
treatment at this program (i.e., number of days clients wait to enter this program)?
1. Outpatient Treatment

2. Residential Treatment

3. Narcotic Treatment Program

____________________
Number of days

____________________
Number of days

__________________________
Number of days

295

4. What has been the average length of time from initial contact to entering treatment for
Prop 36 clients at your program?
_____________Average number of days
5. Does your program offer any services to clients who are awaiting admission to
treatment (e.g. wait listed)?
_______No ______Yes ______ Not applicable (little/no wait time)
6.

If yes, please describe __________________________________

7. If your program offers Narcotic Replacement Therapy (e.g., methadone
maintenance), are strategies to address wait time different for those waiting to receive
NRT compared to those that are not?
_______No ______Yes _____ Not Applicable
8.

If yes, how do they differ? _____________________________________

Does your program have a policy of notifying criminal justice personnel (e.g., court,
probation, parole) if a Prop 36 client is placed on a wait list?
_______No (skip to Question 11) ______Yes
9. Does your program report the probable length of wait time
for treatment entry?
_______No ______Yes
10. Are Prop 36 clients required to pay any fees to your treatment program
(intake fees, for example) before they begin treatment?
____No ____Yes
12-14.What is the typical caseload for a counselor with Prop 36 clients?
(Number of clients per counselor, include Prop 36 and non-Prop 36 clients if applicable)
12. Outpatient Treatment

13. Residential Treatment

______________________
Number of clients per
counselor

______________________
Number of clients per
counselor

296

14. Narcotic Treatment
Program
______________________
Number clients per counselor

SECTION 6. TREATMENT CHARACTERISTICS
No
1. Does your intake process include a motivational interview
designed to assess each Prop 36 client’s readiness for change
and for treatment?
_______

_______

Does your program:
2. Have separate groups based on client level of motivation?

_______

_______

3. Limit the number of unmotivated clients in group?

_______

_______

4. Hold Prop 36 graduation ceremonies?

_______

_______

5. Provide Prop 36 clients with certificates of completion?

_______

_______

6. Provide coupons/gift certificates to clients as motivational
incentives (e.g., for clean urines, treatment compliance)?

_______

_______

7.

To what extent is Motivational Enhancement Therapy used at your program?
Not At
All
0

8.

Yes

Limited
Extent
1

Moderate
Extent
2

Great
Extent
3

Does your program routinely provide a formal standardized orientation session,
explaining processes and obligations associated with the Prop 36 program, to most
Prop 36 clients?
_____No _____Yes
No

Yes

9. Pamphlet/paper…………

_______

_______

10. Video/DVD……………………

_______

_______

11. Verbally…………………………

_______

_______

12. Other………………………………

_______

_______

If yes, how is information usually conveyed:

13. Describe__________________________________
14. Have you heard of the Network for the Improvement
of Addiction Treatment (NIATx)?
_____No (skip to Question 17) _____Yes
Has your treatment program ever
15. Implemented NIATx
procedures?…………………………………………
16. Communicated with representatives from the NIATx
organization?…

297

No

Yes

_______

_______

_______

_______

17. Has your program implemented changes in practices, services, policies, etc., to
improve show rates, reduce treatment drop out, and/or increase retention in the past
year?
_____No _____Yes
If yes, how is the impact of the change typically assessed?
18. Director/management judges the impact based on
observation
19. Changes are discussed at staff meetings
20. Outcome data are systematically collected before and after
the change to measure the effect……………………………
21. Other………………………………………………………

1-3.

No

Yes

_______ _______
_______ _______
_______ _______
_______ _______

22. Please describe: ______________________________
SECTION 7: TREATMENT COMPLETION
What is the average number of days Prop 36 clients are expected to stay in
treatment at this program?

1. Outpatient Treatment

2. Residential Treatment

3. Narcotic Treatment Program

____________________
Number of days

_____________________
Number of days

__________________________
Number of days

4-14. To what extent do the following statements describe reasons why Prop 36 clients
have not completed their planned treatment duration at this program?
Not At
All

Limited
Extent

Moderate
Extent

Great
Extent

4. Unwilling to comply with Prop 36 requirements

0

1

2

3

5. Lack of transportation

0

1

2

3

6. Conflicts with work schedule

0

1

2

3

7. Lack of stable housing

0

1

2

3

8. Family responsibilities

0

1

2

3

9. Probation/parole violation

0

1

2

3

10. Re-arrested

0

1

2

3

11. Relapse

0

1

2

3

12. Insufficient motivation

0

1

2

3

13. Other

0

1

2

3

14.

Please Describe:_____________________________________________

298

15-17. Of the reasons for non-completion listed, what are the top three reasons clients in
your program do not complete treatment?
15. 1st______________________________
16. 2nd _____________________________
17. 3rd _____________________________
18-22. Do you think treatment completion at your program would be improved if Prop
36 clients:
No
Yes
_______ _______

18. Received treatment reminder phone calls
19. Were given more intensive treatment (e.g., more individual or
group sessions) if they were not compiling with treatment
requirements
_______ _______
20. Were given brief jail stays for continued treatment
noncompliance
_______ _______
21. Other
_______ _______
22. Please describe: ________________________________________
ADDITIONAL REMARKS
Is there anything else you would like to tell us regarding the implementation/operation of
Prop 36 in your treatment program? (Attach additional pages if necessary.)

THANK YOU VERY MUCH FOR COMPLETING THIS SURVEY.
Please keep a copy of your completed survey for your records.
Please return this survey in the enclosed postage paid pre-addressed DHL envelope. If
you would like to have DHL pick up your survey, or find a DHL drop off location near
you, call DHL at 1-800-Call-DHL (1-800-225-5345). If you prefer to send your survey
by U.S. mail, please address it to: Joy Yang, UCLA Integrated Substance Abuse
Programs, 1640 S. Sepulveda Blvd., Suite 200, Los Angeles, CA 90025.

299

300

Appendix C: UCLA’s Proposition 36 Focus Groups Information
Ten focus groups were conducted from June through September 2007 to identify promising
and innovative practices from the perspectives and experiences of various stakeholders (e.g.,
treatment providers, county lead agency staff, bench officers, probation department
personnel). Focus groups were held in counties that were diverse in location (Northern,
Central, and Southern California), size (large, medium and small), and setting (urban or
rural). Focus groups with treatment providers included representatives of outpatient drug
free, residential, and narcotic treatment programs.
The groups covered topic areas mutually agreed upon by UCLA and the California
Department of Alcohol and Drug Programs (ADP). Based on their knowledge and
experiences related with these different topic areas, various stakeholder groups were invited
to participate in the focus groups. For example, treatment programs that had participated in
the NIATx process improvement pilot project were interviewed on that topic. To obtain
potentially opposing ends of the spectrum of opinions on Narcotic Treatment Programs
(NTP), UCLA interviewed both a group of NTP providers and a group of bench officers
(judges and commissioners). To gather information on employment practices, UCLA
interviewed one county that had positive Prop 36 client employment outcomes based
CADDS data, and another that had received OTP funds related to employment. However,
although each group may have been selected primarily for their input on specific topics,
wherever time allowed UCLA also took the opportunity to ask all groups questions on all
topics that were relevant to the group.
Table C.1: Focus Group Participants
Focus
Group
Number

No. of Participants

Treatment
Program

Single-County
Stakeholders

1

11

2

4

xx

3

8

xx

4

2

xx

5

13

xx

6

6

xx

7

9

8

10

xx

9

7

xx

10

11

Other
Stake-holder
Group

xx

xx

xx

The tables provides information on the number of participants, their affiliations, and the
topics covered in each focus group. Three focus groups were comprised of representatives
from individual treatment programs, and four were comprised of diverse stakeholders from

301

an individual county. The other three groups were made up of a group of NTP providers,
bench officers, and a group of county administrators from small (minimum base allocation)
counties.
Table C.2: Focus Group Topics
Focus
Group
Number
1

NIATx

2

xx

3

xx

4

xx

Residential

Testing/
Sanctions

Employment

NRT

Mentally Ill/
Homeless

xx

xx

xx

xx

xx

xx
xx

xx

xx

xx

xx

xx
xx

5

xx

xx

xx

6

xx

xx

xx

7

xx

xx

xx

xx

xx

xx

8

xx

xx

xx

xx

xx

9

xx

xx

xx

xx

xx

10

xx

xx

xx

xx

xx

Focus groups were held in private rooms at treatment or county agency facilities, or at
locations chosen by the particular stakeholder group. Each session began with the moderator
and assistant introducing themselves and then welcoming the participants. An informed
consent form describing the purpose, procedures, and confidentiality of the focus group
discussion was given to each participant, reviewed with them, and questions were answered.
Participants were asked to sign the form if they agreed to participate in the research.
Background information was also collected from the participants through a brief pencil-andpaper survey questionnaire that did not include participants’ names. Next, individuals were
invited to select an alias to use during the session. The majority of the focus groups lasted
approximately two hours. An assistant took written notes during the session and a summary
of the discussion was produced afterwards. The focus group discussions were digitally
audio-recorded and later transcribed verbatim by a professional transcription service; the
transcripts were checked against the recording and edited by research staff. Participants
were each paid $25 in cash or money order for their participation if it was not in violation of
county or program policies.
Participants
Eighty-one participants representing treatment (48.1%), county alcohol and drug programs
administration (27.9%), court administration (11.4%), probation (7.6%), local parole (2.5%),
and public defender’s offices (2.5%) participated in the focus groups. Nearly three-quarters
of the participants reported their highest degree earned was a bachelor’s degree or
graduate/professional degree; 12.5% had earned an associate’s degree; 11.3% held a high
school diploma or equivalent; and 2.5% reported another degree (e.g., CSAC). On average,
participants had worked at their current organization for almost 11 years. They also reported
302

working on Prop 36 for an average of 3 years, ranging from less than one month to 6 ½
years. In terms of other background characteristics, the average age of participants was 50
years (range of 24 to 70 years) and the majority of participants were female (59.5%). Most
(67.1%) identified racially/ethnically as white; 12.7% as Hispanic/Latino; 8.9% as
Black/African American; 3.8% as Asian/Pacific Islander; 2.5% as Native American; and
5.0% as multiracial or of another race/ethnicity.
Analysis
Transcript data were coded using Atlas.ti, a qualitative data analysis software package,
according to the pre-determined topic areas, including NIATx process improvement,
employment, residential treatment, drug testing, sanctions, narcotic replacement therapy
treatment, homeless, and mental health; additional codes (e.g., barriers, education, what’s
working) were added to the code list as themes and patterns emerged after reading and
rereading the transcripts. The final code list developed was comprised of 21 primary codes.
These codes aided in identifying discussions relevant to various topics discussed in this
report.

303

304

Appendix to Chapter 1
Appendix 1.1: Prop 36 Eligibility Exceptions
There are some Prop 36 eligibility exceptions. Prop 36 does not apply to any offender
previously convicted of one or more serious or violent felonies, unless the current drug
possession offense occurred after a period of five years in which the offender remained free
of both prison custody and the commission of an offense that resulted in (1) a felony
conviction other than a non-violent drug possession offense or (2) a misdemeanor conviction
involving physical injury or the threat of physical injury to another person. Also ineligible is
any non-violent drug possession offender who has been convicted in the same proceeding of
a misdemeanor not related to the use of drugs or any felony. Prop 36 does not apply to any
offender who, while using a firearm, unlawfully possesses (1) a substance containing cocaine
base, cocaine, heroin, or methamphetamine or (2) a liquid, non-liquid, plant substance, or
hand-rolled cigarette, containing phencyclidine. Prop 36 does not apply to any offender
who, while using a firearm, is unlawfully under the influence of cocaine base, cocaine,
heroin, methamphetamine, or phencyclidine. Prop 36 does not apply to any offender who
refuses drug treatment as a condition of probation or parole.

305

Appendix 1.2: Terms of Proposition 36 Participation for Parolees and Probationers
Factor

Parolees

Controlling Law

Penal Code 1210, 3063.1, 3063.2 Penal Code 1210, 1210.1, 1210.5

Adjudication
Authority
Supervision
Authority
Serious or
Violent
Background

Probationers

Board of Prison Terms
Parole and Community Services
Division, California Department
of Corrections and
Rehabilitation
Parolees who have ever been
convicted of a serious or violent
felony are ineligible.

Disposition of
charges

Superior Court
County probation department

Offenders with prior serious or
violent felony convictions are
eligible if the conviction is more
than five years old and they have
been free of both prison custody
and non-drug possession felony or
violent misdemeanor convictions
during that period.
Original charges remain open for
dismissal upon successful
completion or re-sentencing upon
failure to complete treatment.

Placement in Proposition 36 is
the final disposition. Failure to
complete treatment must be
charged as a new violation.
Term of
Placement on parole occurs
If not already on probation,
supervision
before placement in Proposition
offenders are placed on probation.
36 and will terminate
Probation will not terminate prior
independently of parolees’
to completion of treatment.
progress in treatment.
Disposition of
Parolees become ineligible upon Probationers become ineligible
upon the third violation
drug violations
the second violation subsequent
subsequent to placement (second
to placement (first violation for
violation for those on probation
those on parole before July
before July 2001).
2001).
Source: Joseph Ossmann, Acting Director for the Office of Substance Abuse Programs,
California Department of Corrections and Rehabilitation
Appendix 1.3: Pipeline Analysis
Offenders who choose Prop 36 are referred to assessment and treatment. Assessment entails
a systematic review of the severity of the offender’s drug use and other problems, a decision
regarding appropriate placement in a drug treatment program, and identification of other
service needs. Upon completion of assessment, offenders must report promptly to the
assigned treatment program. Thus, referral is the first step in the Prop 36 pipeline.
Completion of assessment is the second step, and treatment entry is the third.
Information to describe the pipeline was compiled from four sources: the SACPA Reporting
Information System (SRIS) maintained by ADP, the 2007 UCLA Stakeholder Survey, the
306

California Alcohol and Drug Data System (CADDS), and the California Outcome
Monitoring System (CalOMS). The first two of these sources were created specifically for
Prop 36 monitoring and evaluation. The third, CADDS, predates Prop 36, having been
maintained by ADP since July 1991. CalOMS replaced CADDS in 2006.
Each data source had unique value in this analysis but was also subject to limitations. To
overcome these limitations, the pipeline analysis employed a mixture of data taken directly
from these sources along with estimates validated across multiple sources when possible.
ADP’s efforts to validate referral data in 2005-2006 and both referral and assessment data in
2006-2007 provided added confidence in these data elements over that in previous years.
Accordingly, UCLA changed pipeline calculation methodology to maximize use of these
data elements.
2005-2006 Pipeline Methods
Because ADP validated SRIS referral data for the 2005-2006 year by calling county lead
agencies to discuss and confirm or revise the referral counts as necessary, this referral data
was accepted without change, and it was assumed that this figure correctly reflected the
number of unique offenders referred in that year and not the previous year in all counties.
Assessment and placement data were not checked by ADP in the same way for the 20052006 year, however, and therefore the following adjustments and substitutions were used
wherever the available data failed the logic checks described.
•

•

If the reported number of clients assessed exceeded the number referred, the
percentage of referred clients that were assessed in 2006-2007 was applied. For
example if a county referred 100 offenders in 2006-2007 and assessed 75 of
them, the 2005-2006 count of assessed offenders was set at 75% of the 20052006 count of referrals. Since ADP confirmed the counts of assessed offenders
in 2006-2007, this percentage was deemed to be a good approximation of
county practices. This adjustment was used in 6 of the 58 counties. In Los
Angeles County, reliable 2004-2005 numbers were available from a county
report, so this 2004-2005 assessment show rate (85.2%) was averaged with the
2006-2007 show rate (81.5%) to produce a show rate estimate for the
intervening year (2005-2006).
If the reported number of clients placed in treatment exceeded the number
offenders assessed, the number of unique individuals admitted to treatment in
the county through Prop 36 in that year as reported to CADDS, after removing
clients who were admitted in the prior year, was substituted for the treatment
placement count (17 counties). In Los Angeles County, since reliable 20042005 numbers were available from a county report, the percentage of referred
unique offenders who were admitted for that year (70.2%) was applied to the
number of referrals in 2005-2006.

Using CADDS as a data source produces a somewhat conservative estimate for several
reasons. First, CADDS does not include privately funded treatment while counties may
include this in their SRIS counts. Second, while clients admitted in the previous year were
removed to adhere to the definition of placements in the SRIS manual (which instructs

307

counties not to report individuals who entered Prop 36 during the prior reporting period), if
any counties are not removing these counts from their referral and/or assessment numbers,
removing them from the placement count will create an artificially low show rate. Third, the
CADDS unique participant identification variable, which consists of a clients’ first and last
initial, sex, and date of birth, is not entirely unique. Due to the large number of Prop 36
clients, some will share the same identifier, causing an unknown number of clients to be
mistakenly removed during the effort to identify unique clients. Despite this conservative
bias, CADDS may be more reliable than some county-provided estimates, which are
generated by means unknown to UCLA. ADP validation of the referral number should have
caused counties to report this number consistently with SRIS requirements, mitigating the
concern that this figure is being reported with participants from the prior year included.
Finally, as CADDS has been replaced by CalOMS, which uses a different, more specific
identifier, this method will be less vulnerable to problems in the unduplication process.
2006-2007 Pipeline Methods
For the 2006-2007 year, ADP validated SRIS referral and assessment counts by calling
county lead agencies to discuss and confirm the numbers reported. Therefore both counts
were accepted without change. Placement data were not validated in this way, however, and
therefore substitutions were required in cases where the number of clients placed in treatment
exceeded the number of clients assessed. In 11 counties, these placement counts were
replaced by the number of unique clients admitted to Prop 36 treatment in 2006-2007 but not
2005-2006. Use of CalOMS retains the same conservative biases and mitigating
circumstances described above regarding the CADDS data. Although CalOMS has a
superior client identifier, CalOMS data collection began in January 2006. Therefore,
removing clients who entered treatment during the 2005-2006 year requires cross referencing
with CADDS by replicating the CADDS unique identifier. This introduces the problems of
unduplication by the less-specific CADDS identifier described in the preceding section. In
future years, the prior year’s data will also have been collected by the CalOMS system, and
the CADDS identifier will cease to be an issue.

308

Appendix to Chapter 2
Definition of a Treatment Episode
Prop 36 provides up to 365 days of treatment (an additional six months of aftercare
attendance may also be required). Thus, offenders who entered Prop 36 as late as June 30,
2005, (the end of the fourth year) had 365 days in which to complete their Prop 36 treatment
episode. The discharge record for most of them should have appeared in CADDS on or
before June 30, 2006. However, this was not always the case. During the course of their
treatment episode, some clients were transferred from one provider to another. If the transfer
entailed an interruption in treatment, a client’s treatment episode, counting all segments of it,
might have extended beyond one calendar year. Similarly, clients who dropped out of
treatment may have been allowed to re-enter treatment at a later date. They too may have
had a treatment episode of two or more segments spanning more than a calendar year.
UCLA defined the treatment episode as follows: First, clients who entered treatment between
July 1, 2004 and June 30, 2005 were counted as fourth-year Prop 36 clients if their initial
intake record showed a referral from Prop 36 probation or parole. Most Prop 36 clients had
only one treatment segment during that timeframe. Those with two or more segments were
regarded as transfers if the later segment began not more than two days after the earlier
segment ended and even if the intake record for the later segment(s) did not indicate referral
from Prop 36. This procedure maximized the likelihood that the treatment client was still a
Prop 36 participant when the later segment began. It is unlikely that a person could leave
treatment, be dropped from Prop 36, and begin treatment again as a non-Prop 36 client
within such a short window of time. Most transfers occurred within this two-day window (in
a supplemental analysis, the transfer window was extended to 30 days, however, the findings
did not change). Treatment episodes were defined similarly for non-Prop 36 criminal justice
participants and non-criminal justice participants for comparison. Second, to measure time
in treatment, UCLA counted the number of calendar days from intake to discharge for each
segment of the client’s treatment episode. Third, to allow for clients whose time in treatment
may have extended past 365 calendar days (and to allow for lag in data entry as well), UCLA
scanned CADDS for discharges appearing as late as June 2007—two years past the end of
Prop 36’s fourth year. Time in treatment was typically far shorter than 365 days among
offenders who completed their Prop 36 treatment. Hence, an analysis allowing two years for
a discharge to appear in CADDS missed few clients, whether they completed treatment or
dropped out prematurely.
Missing Discharge Data and Completion Rates
The proportion of fourth year clients who had a discharge recorded in CADDS by June 2007
was 92.1%. While it is impossible to know precisely what proportion of the 7.9% missing
discharge records were completions, it is possible to test the sensitivity of the completion
rate calculation to this missing data. If the most extreme negative assumption is made, that
every client without a discharge record did not complete treatment, this has only a very small
effect on the SACPA completion rate, reducing it by 2.8 percentage points. If the opposite
extreme positive assumption is made, that every person with a missing discharge record
completed treatment, which is extremely unlikely, the completion rate would be 5.9
percentage points higher.

309

Among another 10.2% of the records the last discharge status in the treatment episode was
“referred or transferred for further drug/alcohol treatment/recovery.” An unknown portion of
these clients likely did receive further treatment but these subsequent admissions were not
identified either because they were not admitted within the short window of time UCLA
conservatively searched, or because the available client identifiers were insufficient to
identify the new treatment admission as belonging to the same client. While it is impossible
to know what proportion of these referred clients ultimately completed treatment, it is
possible to test the sensitivity of the completion rate calculation to this group. If the most
extreme negative assumption is made, that not one of these clients went on to complete
treatment, this would reduce the overall completion rate only modestly, by 3.6 percentage
points. If the opposite extremely unlikely assumption is made, that all of these clients
completed treatment, the completion rate would be 7.5 percentage points higher.
Realistically, the actual completion rate among these groups is between the positive and
negative extremes described above. A plausible case could be made that it would be closer
to the negative side if missing data tends to occur more often are more often when clients
stop showing up for treatment. Likewise, a plausible argument could be made that a number
of clients with a last discharge of referred/transferred may not have been admitted for further
treatment, and that this could mean that there is a lower completion rate in this group.
However, the calculations above demonstrate that even in these cases UCLA’s completion
rate calculations are not highly sensitive to even the most negative assumptions. Therefore
UCLA has presented completion rates in this report based solely upon the completion data
available without making assumptions regarding these groups.

310

Appendix to Chapter 6
Appendix 6.1: Brief Description of the Treatment System Impact and Outcomes of
Prop 36 (TSI) Study
Led by Yih-Ing Hser, Ph.D. at the UCLA Integrated Substance Abuse Programs, Treatment
System Impact and Outcomes of Prop 36 (TSI) is a NIDA-funded multi-site prospective
treatment outcome study designed to assess the impact of Prop 36 on California’s drug
treatment delivery system and evaluate the effectiveness of services delivered. In 2003,
thirty treatment assessment sites in five counties (Kern, Riverside, Sacramento, San Diego,
San Francisco) were selected for participation based on geographic location, population size,
and diversity of Prop 36 implementation strategy. Two additional counties, Los Angeles and
Shasta, joined the study in 2005.
For TSI’s Treatment Outcome Component, county assessment center or treatment program
staff collected data from all Prop 36 participants assessed for treatment in the selected
counties from November 2003 to December 2006. A sample of the 7,416 participants who
completed the intake assessment was randomly selected for follow-up by telephone with
UCLA-trained interviewers at 3-month and 12-month post assessment. Of 1,588 targeted for
follow-up, 1,464 (92.2%) completed a 3 month interview (another 48 were contacted but not
interviewed because they were unable to respond, were incarcerated, or had died) and 1,290
(81.2%) completed a 12 month interview (another 96 were contacted but were not
interviewed for same reasons as above). A sub-sample of participants also completed an indepth in-person 12 month follow-up interview and provided urine and saliva samples.
Additionally, administrative data was obtained on all participants and included information
on criminal history and mental health services utilization.
For TSI’s System Impact Component, data was collected between 2003 and 2006 via 39
county administrator-level stakeholder surveys and interviews, treatment program surveys
(n=126 in 2003; n=129 in 2005), and focus groups with treatment provider staff (n=37) and
Prop 36 clients (n=50). Topics of interest covered a wide range of subjects, including: Prop
36 implementation planning and design; extent to which Prop 36 implementation reflects
characteristics of drug courts; type of Prop 36-relevant training/information available to
county agency staff; barriers to and facilitators of Prop 36 implementation; processing of
Prop 36 clients; referral networks; diversion options; drug testing; treatment services; staff
workload/burnout; client and outcome data systems; attitudes concerning rehabilitation
verses supervision; and interagency interaction.
For more information on TSI findings, please see:
Fosados, R., Evans, E., & Hser, Y.I. (in press). Ethnic differences in services utilization and
outcomes among Proposition 36 offenders in California. Journal of Substance Abuse
Treatment, 33, 391-399.
Hser, Y.I., Evans, E., Teruya, C., Huang, D., & Anglin, M.D. (2007). Predictors of shortterm treatment outcomes among Proposition 36 clients. Evaluation and Program
Planning, 30, 187-196.

311

Hser, Y.-I., Teruya, C., Brown, A.H., Huang, D., Evans, E., & Anglin, E. (2007). Impact of
California’s Proposition 36 on the drug treatment system: Treatment capacity and
displacement. American Journal of Public Health, 97, 104-109.
Hser, Y.I., Teruya, C., Evans, E.A., Longshore, D., Grella, C., & Farabee, D. (2003).
Treating drug-abusing offenders. Initial findings from a five-county study on the impact
of California’s Proposition 36 on the treatment system and patient outcomes.
Evaluation Review, 27, 479-505.

312

Appendix 6.2: Unemployment Rate by Year, Not Seasonally Adjusted

Statewide
Alameda
Alpine
Amador
Butte
Calaveras
Colusa
Contra Costa
Del Norte
El Dorado
Fresno
Glenn
Humboldt
Imperial
Inyo
Kern
Kings
Lake
Lassen
Los Angeles
Madera
Marin
Mariposa
Mendocino
Merced
Modoc
Mono
Monterey
Napa
Nevada
Orange
Placer
Plumas
Riverside
Sacramento
San Benito

2000
4.9
3.6
6.3
5.2
6.2
5.6
11.5
3.5
7.4
4.1
10.4
8.4
5.8
17.4
4.7
8.2
10.0
7.3
7.1
5.4
8.7
2.8
6.2
5.6
9.6
7.5
4.7
7.4
3.6
4.1
3.5
3.6
7.1
5.4
4.3
6.0

2001
5.4
4.8
7.3
5.1
6.6
5.5
12.8
4.1
8.0
4.3
10.7
8.8
6.0
15.9
4.5
8.6
10.7
7.0
7.3
5.7
9.6
3.5
5.7
5.9
10.1
6.9
4.7
7.8
3.6
4.4
4.0
4.0
7.6
5.5
4.5
6.3

2002
6.7
6.7
7.6
5.8
7.4
6.5
13.8
5.7
8.7
5.2
11.5
9.6
6.7
15.0
5.4
9.8
11.7
8.3
7.7
6.8
10.6
4.9
6.4
6.7
11.0
7.9
5.1
8.9
4.6
5.3
5.0
4.9
8.4
6.5
5.7
8.9

313

2003
6.8
6.9
8.4
6.2
7.6
7.4
14.4
6.1
8.5
5.6
11.7
10.3
6.9
15.6
5.8
10.3
12.0
9.1
7.7
7.0
10.3
4.9
6.9
6.9
11.5
8.7
5.1
9.0
4.9
5.6
4.8
5.1
9.9
6.5
5.9
10.0

2004
6.2
5.9
8.0
5.8
7.4
6.8
13.7
5.4
8.1
5.3
10.5
9.4
6.5
17.1
5.3
9.9
11.0
9.1
7.6
6.5
9.2
4.4
6.7
6.4
10.9
8.8
5.1
8.3
4.8
5.3
4.3
4.8
9.8
6.0
5.6
9.6

2005
5.4
5.1
7.9
5.6
6.8
6.3
12.7
4.9
7.4
4.8
9.0
8.4
6.1
16.0
4.9
8.4
9.5
7.9
8.0
5.3
7.9
3.9
6.4
5.8
10.0
8.0
4.9
7.3
4.4
4.8
3.8
4.3
8.4
5.4
5.0
8.1

2006
4.9
4.4
6.5
5.3
6.2
5.8
12.6
4.3
6.9
4.6
8.0
8.0
5.6
15.3
4.6
7.6
8.5
7.1
8.0
4.7
7.0
3.5
5.6
5.2
9.3
7.7
4.4
7.0
3.9
4.4
3.4
4.2
7.7
5.0
4.7
7.0

San Bernardino
San Diego
San Francisco
San Joaquin
San Luis Obispo
San Mateo
Santa Barbara
Santa Clara
Santa Cruz
Shasta
Sierra
Siskiyou
Solano
Sonoma
Stanislaus
Sutter
Tehama
Trinity
Tulare
Tuolumne
Ventura
Yolo
Yuba

4.8
3.9
3.4
7.0
4.0
2.9
4.4
3.1
5.1
6.1
5.8
7.5
4.6
3.4
7.8
9.4
6.5
9.8
10.4
5.9
4.5
5.0
7.9

5.1
4.2
5.1
7.5
4.0
3.8
4.4
5.1
5.7
6.3
7.5
8.1
4.6
3.7
8.3
9.7
6.5
9.3
11.4
5.9
4.8
5.1
8.5

6.0
5.2
6.9
8.9
4.7
5.7
5.2
8.4
7.4
7.2
8.7
8.9
5.8
5.1
9.7
11.0
7.2
10.1
12.0
6.6
5.8
6.0
9.8

6.3
5.2
6.7
9.2
4.7
5.8
5.1
8.3
7.8
7.6
9.4
9.5
6.3
5.5
9.9
11.2
7.7
10.5
12.3
7.0
5.8
6.4
10.7

5.8
4.7
5.8
8.8
4.6
4.9
4.7
6.4
7.0
7.6
9.7
9.5
5.8
5.0
9.2
10.6
7.4
11.0
11.6
6.9
5.4
6.1
9.7

5.2
4.3
5.0
7.9
4.3
4.3
4.3
5.3
6.3
7.3
8.4
9.0
5.4
4.4
8.4
9.7
6.9
10.2
9.5
6.5
4.8
5.6
9.1

Source: U.S. Bureau of Labor Statistics, www.bls.gov, accessed September 2007

314

4.7
4.0
4.2
7.4
4.0
3.7
4.1
4.5
5.6
6.6
7.5
8.0
4.8
4.0
8.0
8.9
6.5
9.8
8.5
5.9
4.3
5.2
8.8

Appendix to Chapter 7
Change Project Reporting Form
1. PROJECT TITLE
2. What AIM will address?
(choose one, and indicate
baseline and desired goal)

† Reduce waiting time from ______ to ______ days
† Reduce no-shows from ______ to ______ percent each month
† Increase continuation from ______ to ______ percent each month
† Increase admissions from ______ to ______ days

3. LOCATION
4. LEVEL OF CARE
5. What CLIENTS are you
trying to help? (i.e., IOP
clients transferring to
detox, or all IOP clients?
6. CHANGE TEAM
LEADER
7. TEAM MEMBERS
8. Was this change project
spread from another location

___Yes

___ No

From which location? ____________________

9. Indicate any other impacts
(referrals, bed days, etc)
besides the four NIATx
Aims that you intend to
MEASURE as part of this
change project
(Indicate baseline and target
measures - i.e., increase
referrals from 20 to 40 or
reduce time to assessment
from 10 to 4 days)

315

Rapid Cycle
Plan
What is the
Cycle Date
idea/change to be
#
tested?

Do
What change or
action did you
specifically do to
test this
idea/change?

316

Study
Act
What were the
What is the
results? (Please
next step?
include impact on
aim and other
measures)

Project Outcomes (only complete once the project is finished)
1. When did the project stop?
(Enter in mm/dd/yyyy format)
2. What did you LEARN? (e.g.
what were some unexpected
outcomes, lessons learned from
your change efforts)
3. What was the financial impact
of this change (i.e., the
business case for change)?

Sustainability Plan (only complete if you are sustaining the project)
A. Who is the sustain leader?
B. What steps are being
implemented to assure that the
change is sustained?
C. What system is in place to
effectively monitor the
sustainability of the
improvement?
D. At what point would the
change team intervene to get the
project back on track?

317