ARTICLE IN PRESS
Relation of Taser (Electrical Stun Gun) Deployment to Increase
in In-Custody Sudden Deaths
Byron K. Lee, MDa, Eric Vittinghoff, PhDc, Dean Whiteman, BSa, Minna Parka, Linda L. Lau, BSb,
and Zian H. Tseng, MDa,*
Despite controversy concerning their safety, use of electrical stun guns (Tasers) by law
enforcement agencies is increasing. We examined the effect of Taser deployment on rates
of (1) in-custody sudden deaths in the absence of lethal force, (2) lethal force (firearm)
deaths, and (3) officer injuries (OIs) requiring emergency room visits. Under the Public
Records Act and the Freedom of Information Act, 126 police and sheriff departments from
California cities were mailed surveys requesting rates of each of the outcomes of interest for
each of the 5 years preceding Taser deployment through the 5 years after deployment. To
control for population size and crime rates, we used total annual arrests per city as reported
to the Department of Justice. Fifty cities provided predeployment and postdeployment
data on in-custody sudden death, 21 cities reported firearm deaths, and 4 cities reported
OIs. The rate of in-custody sudden death increased 6.4-fold (95% confidence interval
3.2-12.8, p ‫ ؍‬0.006) and the rate of firearm death increased 2.3-fold (95% confidence
interval 1.3– 4.0, p ‫ ؍‬0.003) in the in the first full year after Taser deployment
compared with the average rate in the 5 years before deployment. In years 2 to 5 after
deployment, rates of the 2 events decreased to predeployment levels. We observed no
significant change in the rate of serious OIs after Taser deployment. In conclusion,
although considered by some a safer alternative to firearms, Taser deployment was
associated with a substantial increase in in-custody sudden deaths in the early deployment period, with no decrease in firearm deaths or serious OIs. © 2009 Elsevier Inc.
All rights reserved. (Am J Cardiol 2009;xx:xxx)
Controversy exists as to whether electrical stun guns, also
called neuromuscular incapacitating devices, can cause
sudden death. The most popular brand of neuromuscular
incapacitating devices is the Taser (Taser International,
Scottsdale, Arizona). Tasers eject barbs that deliver a
high-frequency, high-voltage, low-amplitude current to
incapacitate victims by causing momentary skeletal muscle tetany and neuromuscular incapacitation. Although
Tasers are marketed as a safer alternative to subdue
prisoners and suspects in law enforcement custody,1 recent reports have described a temporal association between use of stun guns and Ͼ300 in-custody sudden
deaths in North America.2,3 Despite the possible risks
posed by these devices, Taser deployment by law enforcement agencies continues to grow; they are currently in use
by Ͼ12,000 law enforcement, military, and correctional
agencies in the United States and abroad.4 We sought to

a
Section of Cardiac Electrophysiology, Division of Cardiology, University of California, San Francisco, School of Medicine, San Francisco,
California; bLoyola University Chicago, School of Medicine, Chicago,
Illinois; and cDepartment of Epidemiology and Biostatistics, University of
California, San Francisco, California. Manuscript received October 3,
2008; revised manuscript received and accepted November 18, 2008.
Dr. Tseng is supported by a grant from the National Center for Research Resources, a component of the National Institutes of Health (Bethesda, Maryland), and National Institutes of Health Roadmap for Medical
Research (KL2 RR024130).
*Corresponding author: Tel: 415-476-5706; fax: 415-476-6260.
E-mail address: zhtseng@medicine.ucsf.edu (Z.H. Tseng).

0002-9149/09/$ – see front matter © 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2008.11.046

determine the effect of Taser deployment by law enforcement agencies on rates of (1) in-custody sudden deaths in
the absence of lethal force, (2) lethal force (officer firearmrelated) deaths (LFDs), and (3) serious officer injuries (OIs)
requiring emergency room visits.
Methods
Based on an initial inquiry distributed to police and sheriff
departments in California, 126 were identified as having
recently deployed Tasers. Surveys were then sent to these
departments in a request for data under the Public Records
Act and the Freedom of Information Act. In addition, surveys were sent to police departments of the 10 largest cities
in the United States outside California. We surveyed for
year of Taser deployment and incidence of in-custody sudden deaths in the absence of lethal force in the 5 years
before through 5 years after Taser deployment (years Ϫ5 to
ϩ5). Year 0 was defined as the deployment year, during
which cities implemented Tasers for 1 month to 12 months
of year 0. We defined sudden death on our surveys as
“unexpected in-custody deaths during nonlethal force situations.” Total arrest data combining all felony and misdemeanor arrests for years Ϫ5 through ϩ5 of Taser deployment for all cities surveyed were retrieved from the
California Department of Justice Web site5 and by contacting the Office of the Attorney General directly for more
recent arrest statistics, as necessary. The year 2000 population for each city was obtained from the Census Bureau.6
Departments that returned the first survey were sent a
www.AJConline.org

ARTICLE IN PRESS
2

The American Journal of Cardiology (www.AJConline.org)

Table 1
Population, arrest statistics, and year of Taser deployment of 84
California cities and counties in survey analysis

p=0.003

92,595 Ϯ 164,626
2,921 Ϯ 4,971

4 (4.8%)
3 (3.6%)
46 (55%)
31 (37%)

p=0.001

Events per 100,000 arrests

Characteristic
Population by 2000 census
Total misdemeanor and felony arrests
in year of Taser deployment
Year of Taser deployment
1985–1990
1990–1999
2000–2004
2005–2007

p=0.23

30
25
20
14.1

9.1

10

6.66

5
0
-5
(18)

Values are means Ϯ SDs or numbers of subjects (percentages).

15.1

15

-4
(19)

-3
(21)

-2
(21)

-1
(21)

0
(21)

1
(21)

2
(21)

3
(10)

4
(9)

5
(5)

Years since deployment of Taser
(Number of cities contributing data)
p=0.34

Figure 2. Mean rates of LFDs (by firearms) by year before (years Ϫ5
to Ϫ1, 6.66/100,00 arrests) (dashed line) and after (years 2 to 5, 9.1/
100,000 arrests) (dotted line) Taser deployment.

Events per 100,000 arrests

p=0.006
p=0.73

15

10
5.96

5
1.44

0.93

0
-5
(44)

-4
(49)

-3
(50)

-2
(50)

-1
(50)

0
(47)

1
(40)

2
(50)

3
(29)

4
(19)

5
(9)

Years since deployment of Taser
(Number of cities contributing data)

Figure 1. Mean rates of in-custody sudden deaths in the absence of lethal force
by year before (years Ϫ5 to Ϫ1, 0.93/100,00 arrests) (dashed line) and after
(years 2 to 5, 1.44/100,000 arrests) (dotted line) Taser deployment.

follow-up survey for confirmation of sudden death rates and
for additional data on incidence of LFDs and serious OIs
requiring emergency room visits for the same 11-year period from 5 years before to 5 years after Taser deployment.
Data were compiled in an Excel spreadsheet (Microsoft,
Redmond, Washington) and analyzed using STATA 10
(StatCorp LP, College Station, Texas). To assess secular
trends in sudden death, LFD, and OI, we used Poisson
regression models. Models were run using generalized estimating equations with exchangeable working correlation
and robust SEs to accommodate within-city correlation and
possible overdispersion. The total number of arrests for each
study year was included in the model as an offset. The
sample for each model was restricted to cities that reported
outcomes in the pre- and postdeployment periods, so that
each city serves to some extent as its own control. We used
these models to estimate and compare average outcome
rates in 4 periods: before deployment (years –5 to –1), the
year of deployment (year 0), the first full year after Taser
deployment (year ϩ1), and years 2 to 5 after deployment.
Results
Of the 126 surveys sent to the police and sheriff departments of cities and counties identified as using Tasers, we
received 113 responses (89.7%). We received no completed
surveys from the 10 largest cities in the United States
outside California; 1 city (Detroit, Michigan) was not using

Tasers and the remaining 9 cities declined to release data.
Thirty-two of the original 126 departments declined our
request for data or returned incomplete surveys without data
on sudden death. Thus, this analysis is based on the survey
responses of 84 police and sheriff departments of moderate
to large cities in California that returned survey data on
sudden death. Population, numbers of total felony and misdemeanor arrests, and year of deployment for these 84 cities
are presented in Table 1.
Only 50 of the 84 departments had available data on
sudden death for Ն1 predeployment and 1 postdeployment
year; 34 departments had deployed too recently to report
this information. Rates per 100,000 arrests are summarized
in Figure 1 for the 50 reporting departments. For each city,
year 0 represented a partial-use year, depending on month of
deployment. Over the entire reporting period, we found an
average rate of 1.57 sudden deaths per 100,000 arrests in the
50 cities contributing to the analysis. Using the Poisson
model, we estimate that the rate of sudden death decreased
slightly from 0.93 per 100,000 arrests in the predeployment
period (years Ϫ5 to Ϫ1) to 0.61 per 100,000 arrests in the
deployment year (year 0, p ϭ 0.73). In the first full year
after deployment (year ϩ1), the rate was 5.96 per 100,000
arrests, a 6.4-fold increase (95% confidence interval [CI] 3.2
to 12.8, p ϭ 0.006) over the predeployment period (years
Ϫ5 to Ϫ1). In years 2 to 5 after deployment, the sudden
death rate decreased to 1.44 per 100,000, a significant decrease from the first full year after deployment (year ϩ1, p
ϭ 0.02), but not significantly different from the predeployment period (years Ϫ5 to Ϫ1, p ϭ 0.34).
Thirty-seven police and sheriff departments reported
data on the incidence of LFDs, including 21 providing data
for Ն1 predeployment and 1 postdeployment year. Figure 2
summarizes the LFD findings for these 21 cities. Using the
Poisson model, we found that the rate of LFDs increased
from 6.66 per 100,000 arrests in the predeployment period
(years Ϫ5 to Ϫ1) to 14.1 per 100,000 arrests in the deployment year (year 0), a 2.1-fold increase (95% CI 1.3 to 3.4,
p ϭ 0.001). In the first complete year after deployment
(year ϩ1), the rate of LFDs remained high at 15.1 per
100,000 arrests, a 2.3-fold increase (95% CI 1.3 to 4.0, p ϭ

ARTICLE IN PRESS
Miscellaneous/Tasers and In-Custody Sudden Deaths

p=0.28
p=0.80

Events per 100,000 arrests

p=0.56
1000
800
600
400
200
0
-5
(3)

-4
(4)

-3
(4)

-2
(4)

-1
(4)

0
(4)

1
(4)

2
(4)

3
(1)

4
(1)

Years since deployment of Taser
(Number of cities contributing data)

Figure 3. Rates of serious OI by year before and after Taser deployment.

0.003) over the predeployment period (years Ϫ5 to Ϫ1), but
not significantly higher than the rate in the deployment year
(year 0, p ϭ 0.79). In years 2 to 5 after deployment, the rate
of LFDs decreased to 9.1 per 100,000 arrests, a significant
decrease from the first year after deployment (year ϩ1, p ϭ
0.04) but not significantly different from the predeployment
period (years Ϫ5 to Ϫ1, p ϭ 0.23).
Thirteen police departments returned follow-up surveys
providing data on OIs, but only 4 of these cities provided
data for Ն1 predeployment and 1 postdeployment year.
Figure 3 summarizes the rates of OI per 100,000 arrests for
the 4 departments contributing these data. The rate of OI in
the predeployment period (years Ϫ5 to Ϫ1) was similar
to the OI rates in the year of deployment (year 0), the first
full year after deployment (year ϩ1), and years ϩ2 to ϩ5
after deployment (p ϭ 0.56, 0.80, and 0.28, respectively).
Discussion
In this epidemiologic study of police and sheriff departments of moderate to large cities in California using Tasers,
we found a statistically significant 6.4-fold increase in the
rate of in-custody sudden deaths not involving lethal (firearm) force in the first full year of Taser deployment compared with the predeployment period. Although Taser use
has been advertised to decrease LFDs (by firearms) and
prevent OIs, we observed no decrease in the rate of either
event after Taser deployment. To the contrary, departments
had a twofold increase in the rate of LFDs in the year of
Taser deployment and the first full year after deployment,
whereas the rate of serious OIs requiring visits to an emergency room was unchanged. Rates of sudden deaths and
LFDs decreased to predeployment levels after the first full
year of deployment.
Previous research on the cardiac and physiologic effects
of Tasers have been inconclusive; these studies have mainly
investigated whether Tasers can directly pace the heart into
potentially lethal ventricular tachyarrhythmias by extremely
rapid pacing or discharge during the vulnerable period in the
cardiac cycle.7–11 Anatomic and electrophysiologic differences between humans and pigs in the controlled, fully
anesthetized condition in which these studies were performed limit their generalizability to humans.12 In humans,

3

1 case report has demonstrated capture of ventricular myocardium at high rates,13 and another has described a victim
who was found in ventricular fibrillation after Taser application.14 Other human studies have demonstrated cardiac
safety but were performed with limited Taser applications in
a dorsal position in healthy volunteers at rest.15–18 These
findings are difficult to extrapolate to real-world conditions
in which Tasers are used. Police suspects would be expected
to have unique physiologic (hyperadrenergic state), environmental (restraint techniques, multiple Taser applications
near the heart on the torso), and external (illicit drugs)
influences, any of which may make them more vulnerable to
sudden death.
Some investigators have suggested that Tasers may also
cause sudden death by increasing the risk of excited delirium.19 Excited delirium is a much-debated condition, in
which sudden death occurs after a violent struggle, often
with police officers.20,21 The exact mechanism of excited
delirium is unknown, but it has been speculated that a surge
in adrenergic tone, hyperthermia, or acidosis may decrease
the threshold for life-threatening arrhythmias.19,21 Thus, excited delirium may be another potential mechanism by
which Tasers increase the rate of in-custody sudden death in
the absence of lethal force. The intense pain associated with
Taser applications would certainly lead to an increase in
adrenergic tone that could be a trigger or contributory factor
for excited delirium. Furthermore, studies in animal models
and humans have demonstrated that Taser application can
cause transient acidosis, another potential contributor to
excited delirium.22,23
Notably, we found an increase in sudden deaths and
LFDs in the early period of Taser deployment and then a
decrease in these events to predeployment levels. We speculate that early liberal use of Tasers may have contributed to
these findings, possibly escalating some confrontations to
the point that firearms were necessary. The later decrease in
sudden deaths and LFDs may reflect recognition of the
adverse consequences of Taser application by law enforcement agencies, leading to an adjustment of usage and/or
techniques to result in the observed decrease in the 2 events
to predeployment levels.
Our study has several limitations. First, we did not ask
law enforcement agencies for details about the reported
in-custody deaths, and specifically we did not ask whether
the Taser had actually been applied in those incidents. It is
likely that the Taser was used only in a subset of these
deaths. However, we controlled for non-Taser–related incustody sudden deaths by using each department’s historic
data for comparison. The only common intervention for all
cities studied during this period was deployment of the
Taser. Second, our study is purely observational. Therefore,
we cannot rule out potential confounding by secular crime
or drug-use trends to explain the increase in in-custody
sudden deaths in the first year after deployment. However,
the fact that Tasers were introduced in different years in
each city makes this explanation less plausible. In addition,
our results are based on data from survey responses, which
may have been inaccurate. The responses of the reporting
departments may reflect variable interpretations of the questions, despite instructions that strictly defined in-custody
sudden death, LFD, and serious OI. Fourth, a lack of com-

ARTICLE IN PRESS
4

The American Journal of Cardiology (www.AJConline.org)

plete survey responses from cities led to only a limited
analysis of OIs; thus, this finding may be less reliable.
Likely, the greatest limitation of this study is that the
analysis and results are based on a subset of cities reporting
Taser use. Several California cities and all of the largest
cities in the United States were unwilling to release information regarding Taser use and in-custody sudden deaths.
In our anecdotal experience, several cities with highly publicized Taser-related sudden death events declined to provide data and we speculate that other cities with more
Taser-related deaths may similarly have been less likely to
return our survey. Thus, the observed association of Taser
deployment with an early increase in in-custody sudden
deaths in this study may actually be an underestimate because under-reporting would tend to attenuate any such
association.
Based on this study, further epidemiologic research on
the effect of Taser deployment on real-world outcomes is
warranted. Transparency by law enforcement agencies with
regard to Taser use and in-custody sudden death outcomes
is critical for future studies by independent investigators.
Acknowledgment: We thank Feng Lin, MS, for statistical
assistance with the figures.
1. Law Enforcement Overview, http://www.taser.com/Pages/le_
overview.aspx. Scottsdale, AZ: Taser International, 2008.
2. Amnesty International reiterates call to suspend police use of tasers
following airport death [news release]. http://www.amnesty.org/en/
library/info/AMR20/004/2007/en. Amnesty International, 2008.
3. Amnesty International’s continuing concerns about Taser use. http://
www.amnesty.org/en/library/info/AMR51/030/2006. Amnesty International, 2008.
4. Q4 2007 Investor Presentation. http://www.taser.com/. Scottsdale, AZ:
Taser International, 2007.
5. Statistics, Felony Arrests. http://ag.ca.gov/cjsc/statisticsdatatabs/
ArrestCoFel.php. Office of the Attorney General, 2008.
6. Population Finder. http://www.census.gov/. US Census, 2008.
7. Lakkireddy D, Wallick D, Ryschon K, Chung MK, Butany J, Martin
D, Saliba W, Kowalewski W, Natale A, Tchou PJ. Effects of cocaine
intoxication on the threshold for stun gun induction of ventricular
fibrillation. J Am Coll Cardiol 2006;48:805– 811.

8. Lakkireddy D, Wallick D, Verma A, Ryschon K, Kowalewski W,
Wazni O, Butany J, Martin D, Tchou PJ. Cardiac effects of electrical
stun guns: does position of barbs contact make a difference? Pacing
Clin Electrophysiol 2008;31:398 – 408.
9. Nanthakumar K, Billingsley IM, Masse S, Dorian P, Cameron D,
Chauhan VS, Downar E, Sevaptsidis E. Cardiac electrophysiological
consequences of neuromuscular incapacitating device discharges.
J Am Coll Cardiol 2006;48:798 – 804.
10. Wu JY, Sun H, O’Rourke AP, Huebner S, Rahko PS, Will JA, Webster
JG. Taser dart-to-heart distance that causes ventricular fibrillation in
pigs. IEEE Trans Biomed Eng 2007;54:503–508.
11. Walter RJ, Dennis AJ, Valentino DJ, Margeta B, Nagy KK, Bokhari F,
Wiley DE, Joseph KT, Roberts RR. TASER X26 discharges in swine
produce potentially fatal ventricular arrhythmias. Acad Emerg Med
2008;15:66 –73.
12. Pippin JJ. Taser research in pigs not helpful. J Am Coll Cardiol
2007;49:731–733.
13. Cao M, Shinbane JS, Gillberg JM, Saxon LA. Taser-induced rapid
ventricular myocardial capture demonstrated by pacemaker intracardiac electrograms. J Cardiovasc Electrophysiol 2007;18:876 – 879.
14. Kim PJ, Franklin WH. Ventricular fibrillation after stun-gun discharge.
N Engl J Med 2005;353:958 –959.
15. Levine SD, Sloane CM, Chan TC, Dunford JV, Vilke GM. Cardiac
monitoring of human subjects exposed to the taser. J Emerg Med
2007;33:113–117.
16. Vilke GM, Sloane C, Levine S, Neuman T, Castillo E, Chan TC.
Twelve-lead electrocardiogram monitoring of subjects before and after
voluntary exposure to the Taser X26. Am J Emerg Med 2008;26:1– 4.
17. Sloane CM, Chan TC, Levine SD, Dunford JV, Neuman T, Vilke GM.
Serum troponin I measurement of subjects exposed to the Taser
X-26(R). J Emerg Med 2008;35:29 –32.
18. Vilke GM, Sloane CM, Bouton KD, Kolkhorst FW, Levine SD,
Neuman TS, Castillo EM, Chan TC. Physiological effects of a conducted electrical weapon on human subjects. Ann Emerg Med 2007;
50:569 –575.
19. Strote J, Range Hutson H. Taser use in restraint-related deaths. Prehosp Emerg Care 2006;10:447– 450.
20. Jenkinson E, Neeson C, Bleetman A. The relative risk of police
use-of-force options: evaluating the potential for deployment of electronic weaponry. J Clin Forensic Med 2006;13:229 –241.
21. Stratton SJ, Rogers C, Green K. Sudden death in individuals in hobble
restraints during paramedic transport. Ann Emerg Med 1995;25:710 –
712.
22. Jauchem JR, Sherry CJ, Fines DA, Cook MC. Acidosis, lactate, electrolytes, muscle enzymes, and other factors in the blood of Sus scrofa
following repeated TASER exposures. Forensic Sci Int 2006;161:
20 –30.
23. Ho JD, Miner JR, Lakireddy DR, Bultman LL, Heegaard WG. Cardiovascular and physiologic effects of conducted electrical weapon
discharge in resting adults. Acad Emerg Med 2006;13:589 –595.