Deaths In American Police Custody:

MEDICAL SCHOOL

A12 MONTH SURVEILLANCE STUDY
Jeffrey D. Ho, MD, James R. Miner, MD, William G. Heegaard, MD, MPH, Robert F. Reardon, MD
University of Minnesota Emergency Medicine Program, Department of Emergency Medicine, Hennepin County Medical Center
Minneapolis, Minnesota

UNIVERSITY OF 1111 NESOTA
EMERGENCY MEDICINE PROGRAM

INTRODUCTION

WHO DIES IN CUSTODY

SIGNIFICANT ASSOCIATIONS FOUND

Emergency physicians often care for suspects in police custody.
Sudden and unexpected in-custody death (ICD) of a suspect is an
alarming event that occurs at an unknown rate. There are many
theories of causation, including drug intoxication, excited delirium and
excessive use of force by police. There is currently not an adequate,
universal method in place to track ICD events so it is difficult to know
how often ICD occurs or what factors surround each event. Since ICD
has recently been a popular media topic of human rights groups, this
lack of a universal tracking method may lead to faulty conclusions
regarding causation or association. This report uses a novel and
prospective process to gather ICD data and analyzes the factors
reported for significance.

A review of available media sources reveals the following statistics
about people who die in police custody without trauma.

• Impact Weapons were associated with death in the first 60 minutes
(13114, p=0.019)

Note: Some subjects fit multiple categories, so numbers are not intended to add up to 100%.

• TASER device application was never associated with instantaneous
death (0/50, p=0.001)

METHODS
www.webclippings.com was prospectively queried for American lCD's
over a 12 month period. This data search service scours 1.5 billion
sources daily for requested terms. Data was forwarded for analysis
and included subject gender, age, behavior, arrest force and weapons
used, time of collapse proximal to arrest, and presence of illicit
substance abuse. Custodial agencies were contacted for details as
necessary. Descriptive statistics and chi-square tests for significance
were used.

RESULTS
•
•
•
•

162 lCD's reported
96.3% males
Mean age 35.7 years (SD±9.8, range 15-75).
62.9% exhibited bizarre behavior and 62.3% had confirmed illicit
drug use just prior to arrest.

POLICE FORCE USED:
None 22
Significant Hands-On 111
Intermediate Weapons 84
(Chemical 20, Impact 14, TASER 50, Handcuffs 162)
Deadly 19
TIME TO COLLAPSE:
Instantly 21
Within 1st hour 85
Between 1-48 hours 56

CONCLUSIONS
• ICD in America is largely an event involving males less than 45
years of age engaging in illicit substance abuse

12.3%

8.6%

were shot with chemical spray

were hij wnh impact weapons

.......

...
~

• ICD appears to occur within the first 60 minutes when an intermediate impact weapon is used
were hit with TASER

• ICD appears to never occur instantaneously when a TASER device
is used

"'-

.. ill'-

• This poster represents a preliminary source of the known ICD data
• No current ICD database exists for analysis and we recommend
that professionals from interested disciplines develop universal reporting guidelines to better track these events

JIlt
68.5%

62.3%

100%

went hands-on with officers

ingested illegal drugs

were handcuffed

LIMITATIONS
We recognize that the most significant limitation of this study is the
very point that we are trying to call attention to. The fact that there is
not a current, universal ICD database is concerning. Because we
have relied on the media to report these, it is possible that we have
not included every ICD that has occurred. And, although we have attempted to verify accuracy of each report, it is possible that inaccurate
reporting of force factors and collapse times occurred. We believe
that an opportunity for a national registry of lCD's exists to further
define and study this problem.