The American J ournal of Forensic Medicine and Pathology

©2000 Lippincott W illiams & Wilkins , Inc., Philadelphia

21(1):39- 52, 2000.

Asp:Q.yxial Death During Prone Restraint Revisited
A Report of 21 Cases

Ronald L. O'Halloran.

M.D .•

and Janice G. Frank,

Determining the cause of death when a restrained p erson
suddenly dies Is a problem for death investigators.
Twenty-one cases of death during prone restraint a re reported as examples of the common elements and range
of variation in these apparently asphyxial events. A reasonable diagnosis of restraint asphyxia can usually be
made after ruling out other causes and collec ting supportive participant and witness statements in a timely
fashion. Common elements in this syndrome include
prone restraint with pressure on the upper torso; handcuffing, leg restraint, or hogtying; acute psychosis and
agitation, often stimulant drug induced; physical exertion· and struggle; and obesity. Establis hing a temporal
association between the restraint and the sudden loss of
consciousness/death is critical to making a correct d etermination of cause of death.
Key W ord s: Restraint asphyxia- Positional asphyxiation--Prone restraint- Hogtying-Sudden death in custody- Agitated delirium-Excited delirium-Cocaine-Methamphetamine-Baton- Pepper spray- Taser--Stun
gun.

M .D.

The sudden, unexpected death of an individual
while in police custody is always a matter of public
concern and frequently leads to litigation. Such
high-profile deaths often are a diagnostic dilemma
for medical examiners or coroners and the forensic
p athologists who work with them . T he autopsy
findings are frequently nonspecific, detailed witness
descriptions of the circumstances of the terminal
event are often not initially obtained, and accurate
accounts are difficult to collect later because of potential litigation.
Reports of sudden death of individuals who were
restrained prone, many of whom were also hogtied,
appeared in the 1990s (1 - 5). The tenn hogtying is
used in this paper to refer to the restraint of a person in a prone position with their wrists and ankles
bound together behind the back. Based on such reports, many members of the law enforcement community have discussed the problem of sudden death
during restrain t procedures, and many have attempted to modify or eliminate the use of the hogtied prone position; however, sudden deaths during
prone restraint continue to occur. We present the
previously unreported sudden deaths of 2 1 individuals who died while being restrained in a prone
position and discuss the factors that seem to put
these persons at risk.

METHODS
The case hi stories and autopsy findings of 21
m en who died suddenly while being restrained
prone during the years 1992 to 1996 were reviewed, analyzed, and summarized. In all cases, the
records included interviews with the restrainers and
other witnesses, if any. In most cases, transcripts
of statements or testimony were also reviewed.
Four of the autopsies were done by the authors, and
the case histories were reviewed as an adjunct. The
other cases were seen in consultative review related
to potential litigation.

Manuscript received May I 0, 1999; accepted August 25,
1999.
From the Office of the Medical Examiner, Ventura County,
Ventura, California, U.S.A.
Address correspondence and reprint requests to Ronald L.
O'Halloran, 3291 Lorna Vista Roac' Ventura, CA 93003,
U.S.A.

39

··~

40

)

R. L. 0 'HALLORA N AND J. G. FRANK

CASE RE,PORTS

Case 1
Police responded . to a domestic violence situation. A man in his early thirties was holding his
w ife against her will. He was physically violent and
irrational and did not respond to verbal commands.
Four police officers used control holds and a baton
to force him into a prone position on the floor, hogtied . He seemed to calm down and stop wrestling
shortly before the hogtying was complete d. He was
canied from the residence and laid prone on the
grass in the yard, still hogtied, awaiting transportation to jail. Officers soon noticed that he was unconscious, was not breathing, and had no pulse.
They started cardiopulmonary resuscitation (CPR)
and called' 'paramedics. Paramedics found him asystolic and not breathing. They intubated him and
administered advanced cardiac life suppmi (ACLS),
eventually obtaining an idioventricular rhythm.
Marginal cardiac function was restored, but he remained unconscious without spontaneou s respiration. He was transported to a hospital where he was
maintained on a respirator for 24 hours b efore he
died, having sustained extensive ischemic brain
damage. H is temperature on admission was 36. 1oc
(97°F). He was known to have abused methamphetamine in the recent past and had associated
paranoid ideation.
It was not clear from the investigation how long
the officers stmggled with the man prior to his effo tis subsiding, what pressure was applied to the
prone p erson while restraints were appli ed, or how
long he was prone on the grass before his unconscious state was recognized.
An autopsy disclosed a 175-cm (70-inch), 88.6kg (I 95-lb) man with abrasions and contusions on
his shoulders, anns, and abdomen and n o internal
injuries. The eyes were not described. His heart
weighed 420 g and was normal microscopically.
Blood drawn in the emergency depmiment tested
positive for cocaine and methamphetamine but the
levels were not quantitated. Postmortem blood was
negative for cocaine and contained benzoylecgonine 1.7 mg/L, methamphetamine 0.85 mg/L, and
amphe tamine 0 .08 mg/L. The cause of death was
listed as acute intoxication due to the combined effects of cocaine/methamphetamine/amphetamine,
and the manner of death was listed as accident.
Case 2
A man in his mid-thirties appeared to be intoxicated in public. When approached by police, he
was described as delirious, ai'ated, and sweaty._He
fled on foot and after a long :ursuit was cornered,
Am J Forensic Med Pathol. Vol. 21. No. I. March 2000

and electrical darts (i.e., tasers) were used with little effect. He was eventually hogtied by many police officers and carried to a police car, where he
was placed prone in the back seat. After a few
minutes, he was found not breathing. Paramedics
transported him while asystolic to the hospital,
w here he was pronou nced dead on arrival.
A n autopsy disclosed a 182.5-cm (73-inch),
145.5-kg (320-lb) man w ith many superficial injuries and several da1is in the skin of the trunk. No
petechial hemorrhages were described. No interna l
injuries were found. The heart weighed 480 g, was
di lated, and had microscopic foci of fibrosis. Toxicology blood test results disclosed cocaine 1.9
mg/L and benzolecgonine 2.7 m g/L. The death was
certified as due to cocaine intoxication and asphyxia from restraint, and the m anner of death was
listed as homicide.
Case 3
Police responded to reports of a man in his midforties walking the street, yelling, and hitting obj ects. When officers approached the man , h e
appeared to be under the influence, delusional, and
incoherent. He fled and was pursued on foot. When
caught, he was delirious and combative. Pepper
spray and baton strikes had no apparent effect.
Three to five officers eventually forced him in the
prone position on the ground w ith weight applied
to h is back and legs, and his arms were handcuffed
behind his back. Minutes later, police noticed that
he was not moving or breathing. They statied CPR
and called for an ambulance. Paramedics found no
vital signs and were unable to resuscitate him. They
noted that he was sweaty. He was transported to a
hospital, where he was declared dead on arrival.
An autopsy disclosed a 185-cm (74-inch), 86.4kg (190-lb) man with many external bruises but no
internal injuries. No petechial hemorrhages were
described. He had a history of paranoid schizophrenia. The death was certified as due to excited
delirium due to paranoid schizophrenia, and the
manner of death was listed as natural.
Case 4
Police responded to a man in his early thirties
babbling incoherently on the street. He was arrested
for public intoxication and was transported to j ail
in handcuffs. At the jail he became combative an d
was hobbled (i.e., ankles bound) and transported to
another jail. At the second jail he was screaming
and stmggling, clearly agitated but not delirious.
He was dragged into a safety cell stmggling and
sweating. Several officers held him prone on the
floor, applying body weight to the back, arms , and

RESTRAIN T ASPHYXIA
legs as they switched handcuffs. After a few
minu tes of stmggling, he became limp and made
s noring sounds. Officers soon noticed the prisoner
was not breathing. CPR was starte d, and paramedics were called. He w as transported to an em ergency depa rtment and s hortly thereafter was pronounced dead. No vita l signs w ere detected at any
point after the arrival of the ambulance.
An autopsy disclosed a 182.5-cm (73 -inch),
11 3. 6-kg (250-lb) man with many superficia l abrasions and contusions of the h ead, trunk, and extre mities. Intrathoracic petechial hemorrhages were
noted. The h eart w eighed 500 g but was microscopically nonnal. Toxicology blood tests results
revealed cocai ne 1.2 mg/L, benzolecgonine 6..4 m g/
L , cocaethy lene 0.4 mg/L, and alcohol 0.02%. The
death w as certified as due to alcohol and cocaine
toxicity, and the manner was listed as accident.
Case 5
A man in his early forties, with a history of alcohol and cocaine abuse, was arres ted for burglary
a nd booke d into j ail. He had a his tory of alcohol
withdrawa l, and a nurse ordered that he be s tarted
on Dilantin 2 days prior to his death. On the day
of death h e became violent in his cell, was kicking
the door, and was delirious and s w eating. He w as
then handcu ffed and moved to a safety cell. He was
pl aced prone on the floor and held down by two
officers on his back and legs while they attempted
to remove his clothing and cuffs. After an estimated 2- to 5-minute stmggle, he became limp and
turned blue. Chest compressions were s tarted, and
an ambulance was called. H e was transp orted to a
hospital in full cardiopulmonary arrest. After epi nephrine administration and electric shock, the
heart started beating again, but he never regained
consciou sness and died about 18 hours after the
incident. N o temperature was recorded.
An autopsy disclosed a 175-cm (70-inch), 77.3kg (1 70-lb) man with many abrasions and contusions . Intrathorac ic petechial hemorrhages w ere
noted. The heart weighed 480 g and had septal contrac tion bands. The liver w a s fatty and cirrhotic.
T ox icology tests di sclosed only therapeutic levels
of phenytoin and very low levels of chlordiazep oxide. The d eath was certified as due to hy poxic
encephalopathy following a cardiac arrest due to
p ositional asphyxia. Complications of chronic a lcoholism were considered contributing conditions,
and the manner of death was lis ted as accident.
Case 6
A man in his mid-thirties v- :; seen by security
guards wandering suspiciously between cars in a

41

park ing Jot. Three guards c hased him 400 m be fo re
catching him. He was reported to be combative and
talking gibberish and was w restled to the ground.
The guards held him prone, with on e guard s traddling his back, as they handcuffed him. He was
he ld prone for an estima ted 2 minutes, and then h e
stoppe d stmggling. T wo guards went searching for
a weapon, and when th ey came back they discovered h e was not breathing. They started CPR and
called for an ambulance. Paramedi cs adminis tered
A CL S and transported him to a hospital, where he
was pronounced dead on arrival. H e was asystolic
from the time the paramedics arrive d .
The autopsy disclosed a 170-cm ( 68-inch), 81.8kg ( 180-lb) man with cutaneous a brasions and
small contusions. N o p etechiae were seen . The
postmortem blood cocaine level was 1.1 mg/L. The
death w as certified as due to cocaine intoxication,
and the manner was listed as accident.
Case 7
A man in his early thirties was driving a tmck
erratically and was follo wed by a police officer until the tmck broke down. He fled on foot, and the
officer chased him hundreds of m eters and finally
caught him, pitming him to the ground. Two other
office rs arrived, and pepper spray was u sed with no
apparent effect. After a 10- to 15-minu te stmggle
on the ground, he w as fina lly controlled by hogtying him in a prone position. With in a few minutes
after the officers got off the man th ey noticed he
was not breathing. They s tarte d CPR , and he v omited . Within 10 minutes an ambulance atTived. Paramedics removed the wrist and ankle res traints and
continued resuscitation attemp ts, and the man was
transported to a hospital, w here he was pronounced
dead on arri val. He had a his tory of previous paranoid and psychotic behavior after methamphetamine use.
An autopsy disclosed a 1<?2.5-cm (65-inch),
63.6-kg (140-lb), muscular man. Only injuries of
the wris ts and ankles w ere no ted . Internal organs
were n ot weighed. Toxico logy tests disclosed
methamph etamine 0.1 mg/L in postmortem blood.
The cause of death was listed as due to methamphetam ine tox icity, and the m anner of death w as
mled an accident by a coroner inquest jury.
Case 8
A man in his mid-th irties drove hi s speeding vehicle into a restaurant parking lot, followed by police. Civilians and police observed him acting
agitated and talking to God. He asked for a knife
to kill himself, and later bolted to a police car and
went for a gun. A police officer pulled him from
Am J For e11sic M ed Pathol. Vol. 2 1, No. 1. March 2000

42

R. L. 0 'HALLORAN A ND J. G. FRANK

the car before he could get the gun, but he began
chasing the officer. Several bursts of pepper
sprayed were used, and the man was stmck with a
baton with little effect. Eventua lly, seven police officers wrestled him to the ground, and after a 5- to
10-minute prone struggle, he was controlled with
handcuffs behind his back and his ankles bound.
One officer w as straddling his lower back area, one
held a foreann on his upper back, and more officers
were holding down his upper torso when he lost
consciousness. Officers administered CPR u ntil
paramedics arrived minutes later. Th e man was
without vital signs during tran sport to a hospital,
where he was pronounced dead on arrival. He had
a h istory of a bipolar or chronic schizophrenic psychotic disorder.
An autopsy disclosed a 182.5-cm (73-inch), 127kg (280-lb) man w ith abrasions and contusions of
the head, back, and extremities. Laryngeal mucosal
petechial hemorrhages were seen, along ·w ith faint
neck strap muscle contusions. The h eart was estimated to weigh 450 g. Minimal patchy areas of subendocardial fibrosis were note d. Postmortem toxicology drug screens were negative. The cause of
death was listed as stress induced cardiopulmonary
fai lure due to restraint and acute psychotic episode.
Schizophrenia, blunt force injuries, and cardiac hypertrophy were listed as contributing factors. The
manner of death was listed as accident.

Case 9
A severely mentally retarded teenaged boy was
being disruptive and damag ing property at a public
residential care facility. Two custodial officers
cuffed his hands behind his back as they held him
prone on the floor for an estimated 3 to 10 minutes.
One officer placed a knee on his back and pulled
upward on his wrists while the o ther put pressure
on his upper back or neck area w ith a forearm and
body weight. Some witnesses said a n eck hold may
have been applied. A third officer held his legs.
Several wi tnesses heard a wheezing sound before
he stopped struggling. H e was soon noted to be
unconscious and to have stopped breathing. CPR
attempts by staff and later by paramedics were unsuccessful, and h e was pronounced dead on arrival
at a hospital.
An autopsy disclosed a 167 .5-cm (67-inch),
104.5-kg (230-lb) teenaged boy with abrasions and
contusions of the head, back, arms, and legs. A
neck muscle contus ion and h emorrhage beside the
thyroid cartilage were noted, but no pe techial hemorrhages were seen. Postmortem blood toxicology
tests detected therapeutic · ·vels of doxepin and
thioridazine. The cause of ..eath was listed a s asAm J Forensic Mt•d Pathol. Vol. 21. No. I . March 2000

phyxiation due to restraint, and the m anner of death
was listed as undetennined.

Case 10
A teenaged boy w as arrested for suspected intoxication. H e was combative and appeared delirious. After a stmggle, police transported him to jail
in the hogtied prone p osition in the back of a police
car. At the jail he was placed on the concrete floor,
reportedly lying on his side, still hogtied, stmggling, swearing, sweating, and spitting. Six officers
held him down, and a towel was placed around his
face and neck area to prevent spitting or biting and
control head movement. After approximately 3 to
5 minutes on the fl oor, he stopped mov ing, and - 1
minute later, officers realized he was not conscious
and summoned help. Resuscitation attempts by jail
staff and later by paramedics were unsuccessful ,
and he was pronounced dead after 50 minutes in a
hospital emergency department. He had a 3-year
history of psychosis.
An autopsy disclosed a 190-cm (76-inch), 109kg (240-lb) teenaged boy w ith palpebral conjuncti val petechiae. Scattered contusions and abrasions
of the head, extremities, and tmnk were observed.
The heart weighed 450 g. A haloperidol level within
the therap eutic range was found in the blood. The
cause of death was certified as dysrhythmia due to
positional hypoxia due to excited delirium while
restrained. The manner of death was recorded as
natural.
Case 11
Police were dispatched to a minor traffic accident where they found a man in his mid-thirties
acting bizarre and sweating profusely. He appeared
to be delirious and under the influence of "PCP"
and was arrested afte r a stmggle. He was transported to j a il in the hogtied prone position in the
back seat of a police car. He continued to b e combative and incoherent at the jail and was placed in
a safety cell. While police attempted to remove his
restraints and clothing, he was held prone by six
detention officers. Two officers had their knees on
his back and one was holding his shoulders when
he suddenly relaxed. He was soon noticed to have
stopped breathing. Paramedics were summoned,
but resuscitation attempts were unsuccessful, and
he was pronounced dead at the jail. T he struggle in
the safety cell lasted about 5 minutes.
An autopsy disclosed a 170-cm (68 inch), 8 1.8kg ( 180-lb) man with conjunctival petechial hemorrhages, intramuscular hemorrhages in the neck, a
large contusion on the back, and many contusions
and abrasions of the head, arms, and legs. The heart

RESTRAINT ASPHYXIA
appeared normal and :weighed 4 10 g. P ostmortem
blood contained methamphetamine 1.7 mg/L, and
the urine level of methamphetamine was 34 mg/L.
The cause of death was listed as traumatic asphyxia
with neck compress ion and positional asphyxia.
The manner of death was listed as accident.
Case 12
A severely menta lly retarded boy in his midteens was damaging property in a private residential care facility. Attempts at verbal control were
unsuccessful , and he was wrestled to the floor by
three adu lt male attendants. He was he ld prone with
a pillow under his head for about 10 minutes, with
attendants holding down each arm/shoulder and
one holding his legs, when they noticed he had
stopped struggling and was not breathing. CPR was
initiated. When paramedics arrived, they detected
no vital signs and an idi ovcntricular cardiac
rhythm. At the emergency department only pulseless electrical cardiac activity was present, and he
was pronounced dead within 30 minutes. He had a
history of mental retardation, cerebral palsy, epilepsy, autism, and attention deficit disorder.
An autopsy disclosed a 177 .5 -cm (71-inch),
113.6-kg (250-lb) teenaged boy w ith petechial
hemorrhages of the eyes, pleura, and epicardium.
The heart weighed 410 g. Microscopic foci of fibrosis were seen in the he art. Fluoxetine and its
metaboli te, thioridazine, and mesoridazine were
identified in postm ortem blood in therapeutic concentrations. The cause of death was certified as asphyxia due to restraint and suffocation, w ith
cardiom yopathy listed as a contributing condition.
The manner of death was ruled as accident.
Case 13
A man in his early thirties was in a hospital for
2 days while being treated for hypertension and ren al failure. Because of verbal ab use and physical
threats against the staff, he was physically escorted
by two hospital security s taff to the emergency
area, where a struggle developed. He was he ld
prone on the floor for an estimated 2 or 3 minutes
with weight on his back, his anns pulled upward
behind his back, and his legs held down. He said
that he could not breath several times. His
breathing became s hallow before he lost consciousness. Resusc itation and ACLS measures were initiated promptly. Sinus bradycardia was detected
initially but deteriorated quickly to pulseless electrical activity and death. No pulse was ever palpated.
An autopsy indicated no xtemal or internal injuries in this 167.5-cm (67- .,Jch), 81.8-kg (180-lb)

43

man. The heart weighed 750 g and sho'?!'ed severe
left ventricular hypertrophy. The lungs had emphysematous changes. The kidneys had vascular and
glomerular sclerosis. Toxicology tests w ere negative. The case was submi tted to the grand jury. The
cause of death was lis ted as sudden cardiac death
during restraint due to positional asphyxia and the
manner of death was called hom icide.
Case 14
Police responded to a residentia l domestic abuse
call. A man in his middle f01ties was escorted out
of his residence, at which point he resisted arrest.
A struggle ens ued, p epper spray was used, and he
was eventually hogtied and transported to j ail in
the prone position in the back seat of a police car.
At the j ail he continued to be combative, yelling
obscenities and spitting, so a dust mask was put on
his face. He was caiTied to a holding cell, w here
he was placed in the prone position and s tripsearched. He was held prone for severa l minutes
by several officers, with a disputed amount of
weight placed on the upper torso. While handcuffs
were removed and others applied, he suddenly
stopped resisting, was noted to be not moving, and
was found to have absent vita l signs. Resuscitation
attempts were initiated by a j ail nurse, and paramedics were s ummoned. When they arrived, he
was in ventricular fibri llation which progressed to
asystole during transport to the hospital, where he
was pronounced dead. He had a history of illicit
drug use. Officers stated that during the encounter
he seemed paranoid about his wife and called to
God.
An autopsy of this 170-cm (68-inch), 72. 7-kg
(160-lb) man disclosed eye and oral petechial hemorrhages, a bite on the tongue, lip lacerations, and
scattered abrasions and contusions of the extremities
and tmnk. The heart weighed 450 g and was microscopically normal. Toxicology testing disclosed a
postmortem blood cocaine level of 1.2 mg/L and a
benzoylecgonine level of 3.2 mg/L. The cause of
death was listed as cardiac arrhythm ia due to exertion and cocaine toxicity, wi th restraint asphyxia
listed as a contributing factor. The manner of death
was ruled accidental.
Case 15
Police responded to a domestic violence report
at a residence. T hey confronted a m an in his midforti es who would not respond to verbal commands
and assaulted the officers. Pepper spray was used
with no effect. He was wrestled _to the floor, hit
several times with a flashlight, and eventually controlled in a prone position by officers pinning and
Am J Forensic Mc:d Parhol. Vol. 21. No. 1, March 2000

44

R. L. 0 'HALLORAN AND J. G. FRANK

shackling his legs, cuffing his wrists behind his
back, and holding his chest down· with feet and
knees on his back and near the base of his neck.
About I 0 m inutes later, he was discovered to be
nonresponsive. Resuscitation efforts were initiated
and an ambu lance was called . When paramedics
arrived, he had no vital signs. ACLS m easures were
unsuccessful , and he was pronounced dea d on arrival at a hospita l. He had a psychiatric history of
paranoid schizophrenia fo r several yea rs, associated
w ith poor impulse control , delusions, and homic idal threats. The initial call for po lice assistance
involved delusions about spousa l infidelity.
The autopsy of this 162.5-cm (65-inch), 8 1.8-kg
(180-lb) man disclosed many abrasions and contusions of the face, anns, legs, and trunk. Two linear lacerations were found on the scalp. Bilateral
sc leral hemorrhages, a cutaneous neck contusion
and abrasion, and a hemorrhage in the sternocleidomastoid muscle were seen. Postmortem toxicol ogy tests of blood identified chlorpromazine and
diphenhydramine in therapeutic concentrations.
The cause of death was listed as asphyxia by com pression of the neck and chest due to restraint. The
manner was listed as homicide.

nonexistent beings . Officers fo und and confro nted
him. He became physically a ggressive toward the
o fficers, and a 5-minute str ugg le ensued. Pepper
spray was used with no effect, and he was wrestled
to the ground. After being he ld prone by several
officers fo r an estimated 2 or 3 minutes while being
handcuffed w ith pressure on his waist and upper
body and with his legs held down, officers noticed
he was not m oving. Restraints were immediately
removed, and ambulance p ersonnel at the scene
were summoned. Agonal respiratory efforts quickly
ceased, and the man was transported asystolic to
the hospital, where death was pronounced. Ninety
minutes after death, the rec ta l temperature was
36 .8°C (98.3°F) . He h ad a cocaine delusional disorder 2 years prior to death.
An autopsy of this 180-cm (72-inch), 140.9-kg
(31 0-lb) m an revealed ocu lar petechial hemorrh ages along with abras ions and contusions of hi s
face and extremiti es. The heart weighed 430 g. Autopsy blood contained 5 .4 mg/L of coca ine and 1.3
mg/L of benzoylecgonine. The cause of death w as
certified as anhythmi a/asphyxia while restrained
clue to cocaine toxici ty. The manner was listed as
accident.

Case 16
A man in his mid-twenties wh o had used crack
cocaine began ye lling incoherently about God and
the devil while banging on walls and windows. His
male companion tri ed to quiet him, which led to a
st ruggle. The companion told investigators that he
pinned the man prone on the floor by sitting on his
back , pulling his arms upward behind his back, and
applying pressure to his shoulders. After several
minutes, the man stopped struggling, and his companion got off hi s back. He heard a few short gasps
but never saw the m an move again. The companion
eventually decided the man was dead and hours
later told friends w ho then called police.
An autopsy of th is I 70-cm (68-inch), 63 .6-cm
(1 40-lb) man disclosed bilateral ocular petechiae,
contusions of the lips, and several areas of intem al
neck soft ti ssue hemorrhage with intact laryngeal
stmctures. Other abrasions and contusions of the
trunk and extremi ties were seen. The postmortem
blood levels were cocaine 0.02 mg/L , benzoylecgonine 0.4 mg!L, and ecgonine methyl ester 0.4 mg/
L. The cause of death was listed as manual strangulation, and the manner was listed as homicide.

Case 18
A man in his mid-twenties was seen walking
down the street, talking and gesturing to nonexistent persons . He broke into a house and stole a
k.Jiife . P olice were summoned and found the man
c ra wling on his hands and knees near a p ark where
he had assaul ted someone. Poli ce confronted him,
a struggle occur red, and h e was w restled to the
g round and held supine. Fire personnel noted a
rapid pulse and incoherent sp eech. An ambu lance
ani ved, he was given oxygen, and he eventually
quieted. He suddenly jumpe d up, ran, and was tackled again by pol ice. This time he was held prone
w ith officers ' body weight on his back and legs.
Handcuffs were attache d w ith his anns behind his
back, and hobble restraints were wrapped around
his ankles. He was held down an estimated 2
minutes by three officers until he became quiet and
officers removed their weight. Officials were discussing w hat to do with him when a paramedic
noticed he was not breathing and had no pulse.
Handcuffs were removed, and resuscitation attempts were started. Pulseless electrical activity
(idioventric ular rhytlm1) was initially detected. Despite ACLS measures during transport to the hospital, he became asystolic. He was. pronounced
dead after 30 minutes in the emergency department.
Twenty minutes later, the rectal temperature was
42.2°C (108°F). He had a history of drug use.

Case 17
Police were called because a man in his early
forties had snorted cocai1 and left his home, running, stumbling, and screaming incoherently to
Am J Forensic Mer/ Parhol. Yo/. Zl, No. I . March 2000

RESTRAINT ASPHYXIA
An autopsy disclosed a 180-cm (72-inch), 100kg (220-lb) man with conjunctival- and epicardial
petechial hemorrhages. Multiple contusions and abras ions were found on the head, extremities, and
trunk. A contus ion was found in a sternocleidomastoid muscle. The heart weighed 430 g and was
microscopically nom1al. Postmortem blood contained cocaine 0.23 mg/L and benzoylecgonine 2.3
mg/L. The cause of death was lis ted as asphyxia/
arrhythmia during prone restraint du e to cocaine
induced agitated delirium. Coca ine-induced hyperthem1ia was listed as a contr ibuting conditi on, and
the manner of death was listed as accident.
Case 19
Police were called because of complaints of a
man in his late twenties dam aging property. Heappeared agitated and said he was having a heart attack. Police called for an ambulance, but w hile
waiting the man became verbally aggressive, spoke
inappropriately in religious and sexual te1ms, failed
to respond to commands, and appeared to feign a
seizure. Pepper spray was used without effect. The
man was restrained in the prone position for an
estimated 1 to 5 minutes on a lawn with an officer
straddling his buttocks and another with a knee on
his back while handcuffs were attached a nd his legs
were he ld. Paramedics arrived during the s truggle,
and it was decided to transport the man to the h ospital on a gurney in leather restraints. Shortly before or during the transfer to the gumey, he stopped
breathing. Despite ACLS meas ures, only intermitte nt pulseless electrical activity was detected. He
was pronounced dead in the e mergency department. He had a history of bipolar disorder with
episodes of mania as well a s drug abuse.
An au topsy disclosed a 180-cm (72- inch), 13 1.8kg (290-lb) man with ocular and facial petechial
hemorrhages and contu sions and abrasions of the
face, extremities, and tnmk. The h eart weighed 430
g and was microscopically nonnal. Autopsy blood
contained the following concentrations of drugs:
delta-9-THC 0.028 mg/L , free codeine 0.18 mg!L,
tota l codeine 0.40 mg/L, fenflurami ne 0.05 mg/L,
doxylamine 0.068 mg/L, and trace amounts of dextromethorphan. The cause of death was listed as
card iac anhythmia due to asphyxia during prone
res trai nt. Acute and chronic psychosis, THC intoxication, struggle, obesity and a m a lpositioned endotracheal tube were listed as contributory factors.
Th e manner of death was lis te d a s accident.
Case 20
An obese man in his late t e nties with a history
of epilepsy had a grand mal seizure at the home of

45

a friend . While one friend c rad led hi s head during
the seizure, another fri end called 9 11 . After- the seizure, the man jumped up, beca me combative, and
seemed confused. A third friend arrived, and the
three restrained the m an in the prone position on
the floor by applying a " full ne lson" and lying
across his legs. When the fire depm1ment arrived,
three firefighters assisted in holding the man down
while another called for police assis tan·c e. When
the first police officer arrived, he rested a knee on
the man' s back when applying handcuffs w hil e the
man was held by firefighters and his friends. One
person stated he h eard the m an say ' ' Get off, I
can 't breathe.'' Approxim ately I 2 minutes after the
s tart of the restraint process, more police arrived
and saw the man still stm ggling. Straps were
wrapped around his a nkles, and the man was placed
prone on a s tretcher, where he appeared to lose
strength and had a pulse rate of 120 beats per minute. Within seconds of being held prone on the
stretcher he became calm, s topped breathing, and
had no pulse. Restraints were removed and CPR
commenced while a call for paramedics went out.
Four minutes later, p aramedics started full ACLS
measures. He was defibrillated several times at the
scene and en route to the hospital without ever regaining life signs and was pronounced dead in the
hospital. He had epilepsy since birth, was having
several seizures per m onth, and was taking Dilantin.
An autopsy disclosed a 175-cm (70-inch), 118kg (260-lb) man w ith abras ions on the wrists and
hip but no petechial hemorrhages or other injuries.
The heart weighed 510 g and was microscopically
normal. Toxicologic tests of p ostmortem blood revealed phenytoin 0.2 mg/L and carboxy-THC 0.007
mg/L. The death was certified as cardiac arrhythmia due to agitated de lirium with restra int. Hypertrophic hea11 disease was listed as a contributing
factor. The manner of death was listed as accident.
Case 21
Police were called because a man in his early
f011ies, said to be mentally ill, hysteri ca l, violent,
and having a gun, was dancing in street traffic,
s houting incomprehensibly. When two officers arrived, the man grabbed one officer by the clothing;
the other officer joined in the stmggk, and all three
fell to the pavement. Three more officers ani ved.
Pepper spray was used w ith no effect. Several officers wrestled the man to the ground and held him
in the prone position. One officer had both knees
on his back while handcuffs and ankle cuffs were
applied. Firemen arrived and suggested using a
backboard for control. Multiple officers moved the
man and placed him prone on the backboard. Straps
Am J Forensic /\.fed Parhol. Vol. 2 1, No. I. March 2000

R. L. 0 'HALLORAN AND J. G. FRANK

46

were secured across his pack and legs. They soon
noticed he had ceased struggling ar\d had no pulse.
Paramedics were called. The total prone restraint
time was approximately 8 minutes. When paramedics arrived 4 minutes later, the man was in full
cardiac arrest. They started ACLS but were imable
to restore cardiac function, and he was pronounced
dead at the hospital. He had a history of crack cocaine use with periods of delirious hyperactive behavior. He stopped taking his prescription Thorazine 2 weeks prior to death.
An autopsy disclosed a 167 .5-cm (67-inch),
86.4-kg (190-lb) man with conjunctival petechial
hemorrhages and scattered contusions, abrasions,
and lacerations of his extremities and face. The
heart weighed 340 g and was described as having
a 1-cm-long myocardial band over the left anterior
descending coronary artery. Postmortem blood
contained cocaine 0.2 mg/L and benztropine (Cogentin) 3 ng/ml. The cause of death was listed as
agitated delirium with restraint. Recent cocaine use
and tunnel coronary artery were listed as contributing conditions. The manner was listed as accident.
ANALYSIS
These 21 case reports represent a nonrandom
sample of sudden deaths that occurred during restraint from 1992 to 1996 in the United States,
mostly in California. Except for the 4 deaths that
we autopsied, all cases came to our attention because of litigation. In many of the cases the details
of the circumstance and timing of the sudden Joss
of consciousness were developed during investigation subsequent to death certification.
All of the decedents were male, ranging in age
from 17 to 45 years. All were involuntarily held in
a prone position when they lost consciousness except for case I 0, who was reportedly held on his
side but had earlier been prone. Four were hogtied
at the time they were noticed to be unconscious;
the remainder had body weight applied to the upper
torso at the time they lost consciousness. All who
lost consciousness while hogtied had weight placed
on their upper torso during the hogtying process.
Eighteen persons were handcuffed behind the back;
the other 3 had their arms restrained manually.
Eleven had ankle or lower leg restraints; the remainder had body weight applied to the legs. Two
were reported to have said that they could not
breathe prior to dying. The number of persons applying restraint ranged from one to seven. Best estimates of the time held pror- ~ ranged from 2 to 12
minutes. Fifteen of the 21 in~ H.lents involved police
Am J Forensic Med ?athol. Vol. 21. No. I. March 2000

only; 3 involved private security or custodial officers; 2 involved lay persons, and 1 involved both
police and firefighters. Five of the incidents occurred in a jail or detention center; 2 occurred in
health care facilities. Seven of the restraint episodes
were preceded by the use of pepper spray alone, 3
by use of pepper spray and a baton, I by baton use
alone, and I by use of a taser. In 2 instances a towel
or a mask was on the persons face during restraint.
In all but 1 case extensive resuscitation efforts
were made, with initial cardiopulmonary resuscitation followed by advanced life-support measures,
including intubation. All but 2 persons were taken
to hospitals, where they were pronounced dead.
Eight of the decedents had a history of chronic
mental illness excluding substance abuse. Eight had
a history of substance abuse. Seventeen appeared
to be acutely delirious. Eleven had stimulant drugs
found in their blood at autopsy (8 cocaine, 2 methamphetamine, 1 both cocaine and methamphetamine). In the 8 who had cocaine found in postmortem blood, concentrations ranged from 0.02 to
5.4 mg/L (mean, 1.4 mg/L). Postmortem temperatures were taken in only 3 cases, and in only 1 was
hyperthem1ia detected. Six were noted to be sweaty
prior to death.
Six decedents could be considered obese, having
a body mass index (BMI) > 30; 9 were overweight
(BMI = 25- 30), and 6 were of normal weight
(BMI < 25). In 5 autopsies, heart abnom1alities
were described. In 2 cases the heart weight was
more than 2 standard deviations above the predicted mean for males, adjusted for body weight
(6). In 4 cases the heart weight was between 1 and
2 standard deviations above the mean. In 3 cases
the heart was not weighed but was reported as normal. Petechial hemorrhages were described in the
eyes in 10 cases and were noted in the thorax in 2
additional cases. Evidence of soft tissue injury
(hemonhage) in the neck was noted in 5 cases, 4
of which also had ocular petechiae.
Death certificates listed asphyxia or a similar
tem1 in the cause of death section in 13 cases, and
8 listed dmgs of some so1i. The manner of death
was listed as accident in 14 cases, homicide in 4
cases, natural in 2 cases, and undetermined in 1
case. Selected circumstances of the deaths and the
causes and manners of deaths as reported by the
medical examiner or coroner are summarized in
Table 1.
DISCUSSION
The concept that the sudden death of individuals
held prone during police restraint might be due to

RESTRAINT ASPHYXIA
TABLE 1.

Case history summary

Prone restraint

Behavior

1
2
3
4
5
6
7
8
9
10

Hogtied
Hogtied
Body weighVcuffs
Body weighVcuffs
Body weighVcuffs
Body weighVcuffs
Hogtied
Body weight/cuffs
Body weight/cuffs
Hogtied on side

Paranoid/irrational
Delirium
Delusional/paranoid
Delirium
Delirium/withdrawal
Delirium
Irrational
Delusions
T emper tant ru m
Delirium

Cocaine/methamphetamine
Cocaine
Schizophrenia
Cocaine
Alcohol
Cocaine
Methamphetamine
Schizophrenia
Mental retardation
Psychosis

11
12
13
14
15

Body weight/cuffs
Body weight
Body weight
Body weight/culls
Body weight/cuffs

Delirium
Temper tantrum
Anger/obnoxious
Paranoid delusions
Delirium

Methamphetamine
Mental retardation
Personality trait
Cocaine
Schizophrenia

16
17
18
19
20
21

Body
Body
Body
Body
Body
Body

Delirium
Delirium
Delirium
Delirium
Combative/confused
Delirium

Cocaine
Cocaine
Cocaine
Psychosis/marijuana
Epileptic seizure
Cocaine

weight
weight/cuffs
weight/cuffs
weight/cuffs
weight/cuffs
weighVcuffs

47

Cause of behavior

Cause of death

Manner of death

Drug toxicity
Drugs/asphyxia
Excited delirium/Schizophrenia
Drug toxicity
Positional asphyxia
Drug toxicity
Drug toxicity
Restraint/psychosis
Restraint asphyxia
Positional hypoxia/rest raint/
delirium
Choking/positional asphyxia
Asphyxia/restraint
Positional restraint
Drug toxicity
Asphyxia/chesVneck
compression
Strangulation
Cocaine/restraint
Cocaine/restraint
Restraint asphyxia
Delirium/restraint
Delirium/restraint

Accident
Homicide
Natural
Accident
Accident
Accident
Accident
Accident
Undetermined
Natural

asphyxia, even though neck holds were not applied,
is relatively recent. In 1985, W etli and Fishbain
reported 7 sudden deaths during cocaine-induced
p sychosis, some of which occurred while the persons were restrained in police custody (7). The
manner of restraint was not specified, and the
deaths were attributed to cocaine. In 1992, Reay et
al. reported the deaths of 3 men who died while
restrained hogticd in the prone position in the back
seat of police cars and attributed their deaths to
positional asphyxia (1 ). In the same year, the San
Di ego Police Department c irculated a task force report on 7 in-custody deaths; 3 of these persons
were hogtied (2). In 1993, O 'Hallora n and Lewm an
reported 11 delirious m en who died while they
were restrained in a prone position; 9 were hogtied
(3). In 1995, Stratton et al. reported 2 deaths in
hogtied prone patients in ambulances (4). In 1998,
Pollanen et al. rep01ted 21 excited delirium- re lated
restraint deaths betwee n 1988 and 1995 in Ontario,
Canada (5 ); 18 were prone and the other 3 had neck
compression. A lso in I 998, Ross reporte d on factors associated with excited delirium deaths in p olice custody from rep01ts of 61 deaths in various
police agencies in the United States (8) .
Based on concerns raised by reports like these
and kn owledge of other unpublished incidents of
sudden death in custody during prone restraint, articles discussing the "sudden in-custody death syndrome," " hogtying," "positional asphyxia," and
"excited delirium" deaths appeared in the law enforcement literature (2,9, 1 0). P vate companies began promoting and providing products and training

Accident
Accident
Homicide
Accident
Homicide
Homicide
Accident
Accident
Accident
Accident
Accident

to law enforcement agencies addressing the risks of
hogtying, positional asphyxia, and sudden in-custody deaths in the mid- I 990s (1 1).
It is not a new concept that a person can die from
the application of body weight to the thorax.
" Burking," a form of m echanical asphyxia combined w ith smothering which involved sitting on a
person 's chest, was used by the nineteenth century
murderers-for-profit Burke and Hare. D eaths from
asphyxia in individuals knocked down and pinned
by the weight of people on top of them during
crowd stampedes and "huma n pile" situations are
w idely recognized (12) . That homicide can occur
by knee ling or sitting on the back of a prone victim
or suspect, or by hogtying, has been acknowledged
in a m ajor forensic pathology text ( 13). The term
''restraint asphyxia'' has been suggested for asphyxial deaths that occur through ·interference with the
mechanical bellows action of the chest, such as in
the prone hogtied position or in the prone p osition
with arms and legs restrained and weig ht applied
to the back (3, 14). However, alternative explanations for sudden death are frequently offered when
prone restraint while in custody is involved. These
include b lun t force head injury, cardiac an·est from
drug toxicity, acute exhaustive mania/excited delir. ium, e lectrical shock from stun guns, respiratory
arrest from pepper spray, and cardiomyopathy. One
or more of these factors was present in most of the
2 I cases currently reported.
Four of the 21 currently reported cases received
baton blows. Two involved blows to the head with
discernible skin lesions but no skull fractures or
Am J Forensic Med Pa1hol. Vol. 21. No. 1. March 2000

48

R. L. 0 'HALLORAN AND J. G. FRANK

brain injuries at autopsy. In all cases there was a
several-minute period of purposeful activity and
voca lization between the delivery of the baton
s trikes and the later loss of consciousness while
held prone. It seems reasonable to exclude brain
injury as a cause of death when loss of consciousness does not follow the blows to the head within
seconds, when the autopsy discloses no skull fractures or brain injuries, and w h en there are other
reasonable explanations for the death. In 1994,
M irchandani et al. emphasized this point with 4
case reporis of sudden death during poli ce restraint
following a stmggle in men with cocaine-induced
agitated deli riu m who had sustained minor head injuries (15). They suggested cocaine-induced cardiac arrest or mental s tress- related "stress cardiomyopathy" may have caused the deaths. They may
not have considered mechanical asphyxia by chest
compression during the restraint process (restraint
asphyxia) as a possible cause for th e deaths.
Karch has recently summarized the pathologic
effects of cocaine and emphasizes the importance
of considering the physical findings, history, and
scene investigation before attributing a death to cocaine effect ( 16). Cocaine can cause sudden death
and can cause delirium. Blood concentrations in individuals w hose death was attributed to cocaine
overlap those in which cocaine was an incidental
finding and have a wide range (17). Some of the
postulated mechanisms by which cocaine could
cause s udden death without leaving les ions vis ibl e
at autopsy include cardiac arrhythmia or coronary
spasm with myocardial ischemia from catecholamine excess, nonspecific myocardial disease from
chronic ischemia, cardiac electrical conduction
slowing from anesthetic effect in high doses, hyperthennia, and seizures. Methamphetamine is postulated to have similar effects ( 18). Eight of the
cases reported herein had cocaine in their blood at
the time that they lost consciousness while being
restrained, and 3 had m ethamphetam ine, and 1 had
both methamphetamine and cocaine.
So-called ''cocaine delirium '' deaths have been
reported to occur w hile in police custody much
more commonly than other cocaine toxicity deaths
(7,15,16, 19,20). The stress of the struggle has been
hypothesized to be a critical factor precipitating a
lethal cardiac arrhythmia in these police encounters, but considerations of the timing of the lethal
event coincident w ith restraint a nd resp iratory compromise were not addressed. Although it is reasonable to attribute the disturbed mental state of
excited delirium to cocaine or methamphetamine
toxicity in this subset, it WC' ld seem that asphyxia

Am J Fon.:n.<ic A/ed Pa thol. Vol. 2 1. No. I , March 2000

wou ld be the likely immediate cause of death, because the sudden collapse that resulted in death occurred while the person was held in a position that
wo.uld compromise breathing. Stimulant drug toxicity could be considered a contributory cause, because the dmg probably precipitated the abenant
behavior that lead to th e encounter that resulted in
death, and may also have sensitized the heart to an
anhythmia.
"Excited (agitated) d e lirium," loosely defined as
a condition of extreme mental and motor excitem ent w ith confused and unconn ected thoughts,
could be interpreted as present in all but 3 of the
21 cases based on reports of behavior during the
incident and knowledge of prior drug-induced or
psychosis-associated deliri~us or delusional s tates.
Others have postulated that s udden deaths in delirious individuals w ho u sed cocaine chronically are
due to a cocaine-induced brain disord er similar to
neuroleptic malignant syndrome (NMS), with abnormalities in the synaptic concentration and metabolism of the neurotransmitter dopamine proposed
as mechanism s ( 19-22). Similarly, in hospitalized,
agitated p sychiatric patients who may or may not
be taking phenothiazine m edications, sudden deaths
have been reported without significant autopsy
findings to explain the cause (23). Various ly known
as acute exhaustive mania, Bell's mania, lethal catatonia, or acute exhaustive psychosis, some postulate that such patient deaths are related to cardiac
arrhythmias from catecholamine-mediated emotional stress, but the relation to restraint is unclear.
NMS is defined as neuroleptic drug- induced hyperthem1ia with muscle rigidity and is thought to
be produced by disruption of the dopamine-dependent them10regulatory centers in the hypothalamus
and basal ganglia. D eath rates attributed to NMS
have dropped from an estimated 30% of diagnosed
cases to nearly zero with treatment. Autopsy findings are often minimal (24).
Two of the cases reported here had neuroleptic
drugs present in thei r b lood. Eight had medical histories of major psychoses. In 3 cases, the postmortem temperature was re corded, and in 1 case it was
elevated. In 6 of the cases, sweating was noted.
Interestingly, sweating was not noted in the 1 case
with documented hyperthem1ia. Of course, perspiration is a normal physiologic response to vigorous
physical ac tivity, psychologic stresses, and wam1
e nvironments; the firs t two factors were present in
all 21 cases. Sweating does not equate with hyperthermia. Medical examiners and coroners could
help c larify the role, if any, of hyperthermia in
prone restraint deaths by promptly obtaining a post-

RESTRAINT ASPHYXIA
mortem temperature. None of the 2 1 death investigations indicated muscle rigidity. Given the temporal association of the restra int process to the
terminal loss of consciousness in all 2 1 of the repoiied cases, it would seem reasonable to attribute
these deaths to asphyxia during restraint rather than
to agitated de lirium . The agitated delirium and its
associated s tresses, whethe r or not drug induced,
could reasonably be considered predisp osing or
contributing to death.
Chronic mental illness was present in 8 of the
2 I deaths reported and was the probable explanation for these persons' agitated behavior. Three
were schizophrenic, 1 had bipolar disorder, 1 had
both diagnoses, I had undifferentiated psychosis, I
was mentally retarded, and I h ad cerebral palsy
with autism. Two of these persons also had cocaine
in their blood a t autopsy. Obviously, not all cases
of agitation and delirium are cocaine induced.
Pepper spray conta ining oleoresin capsicum has
been implicated in deaths in Califomia by the
American Civil Liberties Union (25). However, reviews of deaths where pepper spray was u sed fail
to reveal convincing evidence of lethality (14,26).
One case report of a custody death attributed to
pepper spray indicated that the victim stopped
struggling while handcuffed and being held prone
(27), suggesting restraint asphyxia. Pepper spray
was used in 7 of the 2 I cases reported herein. In
all cases, witnesses reported n o significant effect
caused by the s pray. In all cases, the spraying was
followed by minutes of voluntary physical acti~ity
and verba lization before loss of consciousness during restra int. In no cases were symptoms of respiratory difficulty following the spraying described,
and in no cases were inflammatory changes of the
respiratory mucosa noted a t autopsy.
Electrical shocking devices (s tun guns and tasers) intended to immobilize people are frequently
used by law enforcement officers, and stun guns
are legally available to the public for self-defense
in many areas. They are generally not considered
lethal weapons, and the fe w reports of fatalities associated with their use have identified other more
likely causes o f death (3 ,28-31 ). A tascr was u sed
in 1 of our reported cases. Several minutes of purposeful activity and verbaliz ation followed the
shocks and preceded the death during restrai nt.
Significant physical exertion was present in most
cases before restraint and in all cases during restraint. Officials involved in res training these people often described the persons as unusually strong
and persis tent in their struggle. Increased oxygen
demand from such physical . ::tivity could increase

49

susceptibility to asphyxiation during resH:aint with
pressure on the chest.
Asys tole was the presenting cardiac arrhythmia
found in 15 of the 20 cases reported that had parame dic response. Five others presented w ith agonal
rhythms, described as pulseless e lectrical activity
in 3, and idioventricular rhythm and fine ventricular fibrillation in the other 2 cases. Pree~isting heart
disease was considered by the certifier of death as
a contributory factor in 4 of the 2 1 cases. Microscopic fibrosis was seen in 3 hearts, and 1 heart
was listed as having hypertrophic cardi om yopathy.
In both cases that had hearts more than 2 standard
deviations above the body weight- and gender-adjusted mean (6), the cardiomegaly was not mentioned on the death certificate.
Obesity has been m entioned as a possible risk
factor for death from positional asphyxia during
hogtying (I ,9). Obesity could contribute to asphyxia when excessive body w eight makes chest
wall movement more difficult wh ile prone and
w hen excessive abdominal fat limits diaphragmatic
motion (14). Similarly, obesity has been thought to
be a possible risk factor for death during cocaineinduced excited delirium by contributing to body
insulation and predisposing to hyperthermia (19).
Fifteen of the 21 cases reported were overweight
or obese, defined as having a BMI (body weight in
kilograms/height in meters squared) greater than 25
(32), supporting an association between obesity and
these deaths.
Other p ossible mechanisms of asphyxia in addition to res tricted breathing from chest/abdominal
com pression were present in several cases. Pathologis ts described the presence of soft tissue hemorrhage in the neck musculature in 5 cases,
suggesting the possibility of neck compression during the restraint process. In 2 of those 5 cases, the
e vidence was compelling enough for " neck compress ion" or " strangulation " to be mentioned on
the death certificate. However, in no case did the
participants or w itnesses describe neck holds. In 1
case w ith neck hemorrhage, a towel was held
around the face, and in another case, a mask was
placed over the nose and mouth. All but 1 p erson
received extensi ve resuscitation efforts, including
intubation, possibly expl aining some of the injuries
noted. In a series of 50 patients intubated in the
field by paramedics, 14% had incidenta l hemorrha ges in the strap muscles of the neck found at
autopsy (33).
Ten of the 21 cases had ocular petechial hemorrhages described at autopsy, and 2 more had intrathoracic petechial hemorrhages which were con-

Am J For£>usic Med Pathol. Vo l. 2 1. No. I. March 2000

R. L. 0 'HALLORAN AND J. G. FRANK

50

sidered inconsistent and donspecific indicators of
asphyxia (34). Two other persons ptoclaimed that
they could not breathe shortly before they lost conscious ness and died. These observations support an
asphyxial mechanism of death.
Asphyxia or a similar term was used on the death
certificate in only 13 of the 21 cases, indicating the
difficulty in recognizing restraint-associated deaths
and the difficulty in diagnosing asphyx ia through
autopsy findings alone. In many cases the m edical
examiner or coroner did not have detailed statements from witnesses or dispatch logs that established the association between the restraint and the
loss of consciousness that lead to death. In all· cases
in which the persons were not fully hogtied when
they lost consciousness, weight was being applied
to the chest area when movement ceased and loss
of consciousness was noticed. In the 4 cases that
were hogticd when loss of consciousness was noticed and in the I person who was restrained lying
on his side, the persons had been restrained with
weight on their backs while held prone moments
before. No statements by witnesses indicated conscious activity by any of the persons following
completion of the hogtying, suggesting that loss of
consciousness may have occuned during the application of restra ints. Table 1 indicates the wide
range of cause of death designations given for these
remarkably similar deaths.
The tenn "restraint asphyxia (asphyxiation) "
was first proposed in 1993 to refer to the sudden
deaths of people who were hogtied or restrained in
the prone position with weight on their backs in
which the evidence suggests an asphyxial death (3).
Previously, in cases involving deaths di scovered
during h ogtying by police , the tem1 "positional
asphyxia" had been used (1). "Mechanical asphyxia" or "traumatic asphyxia " are other accepted terms that could have been used. In the field
of forensic pathology, positional asphyxia has usually referred to deaths in which a victim, often
compromised by alcohol or drugs, cannot escape
from a p osition that inhibits pulmonary gas exchange and in which other causes of death have
been excluded by a thorough autopsy (35). Use of
the tenn ''positional asphyxia'' has lead to som e
confus ion when applied to deaths that occur in hogtied p ersons, because positional asphyxia has usually implied accidental, passive entrapment. The
same confusion applies to deaths that occur from
asphyxia produced by other people restricting a
person's ability to breathe during restraint. When
the nose and mouth are blocked, "suffocation" or
"smothering " are accepte. tem1s. When the neck
is compressed, "choking" or "strangulation" are
Am J Forensic

M~d

Pathol. Vol. 21. No. I, March 2000

commonly used. Reay has discussed hogtying and
prone restraint deaths wi th chest compression in the
context of law enforcement take-downs, refening
to them as restraint asphyxia, and has also discussed the biomechanics of such deaths (1 4) . A recent court decision rested, in part, on a confusing
interplay .between the terms " hog tie" and "positional asphyxia" (36) . We suggest that the tenn
"restraint asphyxia" be used to describe deaths
during restraint that appear to be the result of chest
compression or hogtying. Alternatively, more descriptive tenns such as ''asphyx ia by chest compression" or "asphyx ia during hogtying" could be
used. The major advantage of using a sing le term
for death certification is the increased accuracy in
death certificate coding and vital statistics-based
research.
The manner of death designati ons in these 2 1
cases included 14 accident, 4 homicide, 2 natural,
and 1 undetennined. Given the variation in tem1s
used to describe the causes of these deaths, it is not
surprising to see variation in th e manner of death
also. Because restraint asphyxia deaths are "deaths
at the hands of another,'' it has been argued that
they should be considered homicides (37). Conversely, it has been argued that because death by
res traint asphyxia was not recognized until recently
and information regarding its potential lethality has
not b een circulated widely, it is reasonable to classify them as accidental deaths (3,38). Considering
the amount of discussion during the past decade in
th e forensic pathology, emergency medicine, and
law enforcement literature regarding the risks of
death during hogtying, the argument for class ification as accident becomes weaker. Little has been
written in first-responder literature regarding the
danger of death in delirious persons restrained in
the prone position w ithout the use of arms or legs
for support and with sustained pressure applied to
the back- the circumstance in most of the cases
reported herein. Finally, i( one believes sudden
death dnring agitated delirium is a natural consequence of endogenous psychiatric illness, and if no
res traint asphyxial component of the death is recognized, then a determination of death by natural
causes is understandable.
The magnitude of the problem of sudden death
during prone restraint in the United States remains
unclear. In Ventura C ounty, California, which has
a mostly suburban population averaging 720,000
persons, 8 deaths from restraint asphyxia have occuned during the last 14 years. This translates to a
rate of 0.8 deaths/million/year. This would extrapolate to > 200 deathsllt'!illistJyear in the United
States, but the numbers are too low and the popu-

r

RESTRAINT ASPHYXIA
lation too restricted for more than .a crude projection. A s tudy from the province of Ontario, Canada
(5), which had an average population of 11 million
during the years 1988 to 1995, identified 18 cases
of sudden death during prone restraint in excited
delirious p ersons, producing a rate of 0.2 deaths/
million/year. We know of no reports regarding the
frequency of death during prone maximal restraint,
with or without actual hogtying, in any law enforcement population.
CONCLUSIONS
Despite efforts by law enforcement agencies to
limit hogtying, asphyxial deaths still occur when
suspects are held prone w ith their arms and legs
restrained and weight applied to their backs for
minutes. The tem1 "restraint asphyxia" is proposed for such asphyxial deaths involving prone
restraint and/or hogtying. Such deaths are not
unique to Jaw enforcement. Persons with mental
disorders, especially drug-induced or psychotic illness- induced agitated delirious states, seem to be
at greater risk. It is not clear whether the delirious
state itself or its tendency to precipitate violent encounters with police put them at risk. P hysical exhaustion, preexisting heart disease, and obesity may
also increase risk of death in this situation .
As is the case with many other fom1s of asphyxial death, the autopsy findings in restraint asphyxia can be subtle and nonspecific. Each case
must be evaluated on its own merits and altemative
explanatio ns for the death considered. Accurate diagnosis depends on both a thorough autopsy and a
thorough investigation of the circumstances of the
death. Pointed interviews with witnesses and participants in the restraint, focus ing on the mechanics
of the restraint, the length of time involved, and the
moment when loss of consciousness occurred,
should be conducted soon after the event while
memories are still fresh. Identifying the timing of
the sudden loss of consciousness whil e the p erson
was restrained in a position that compromises the
ability to breathe is essential for establishing a
cause-and-effect relation between restraint and
death.
REFERENCES
I . Reay DT, Fligner CL, Stilwell AD, Amold J . Positional
asphyxia during law enforcement transport. Am J Forensic
Med Pathol 1992; 13:90.
2. Burgreen B, Krosch C, Binkerd V, B1ackboume B . Final
report of the custody death task force. San Diego: San
Diego Police Department, 1992 .
3. O'Halloran RL, Lewman LV. Restraint asphyxiation in
excited delirium. Am J Forensic Med Patho/1993; 14:289.

51

4. Stratton SJ, Rodgers C, Green K. Sudden death in individuals in hobble restraints d uring paramedic transport.
Ann Emcrg Mcd 1995;25:710.
5 . Pollanen MS. C hiasson DA, Cairns JT, Young JG. Unexpected death related to restraint for excited deli rium: a
re trospec tive study of deaths in police custody and in the
community. Can Med Assoc J 1998; 158:1603.
6. Kitzman OW, Scholz DG, Hagen PT. l lstrup OM, Edwards WD. Age-related c hanges in nonnal human hea rts
during the first 10 decades of life. Part II (mantrity): a
quantitati ve anatomic study of 765 specimens fro m subjects 20 to 99 years o ld. Mayo Clin Proc 1988;63:137.
7. Wetli CV, Fishbain D A. Cocaine-induced psychosis and
sudden death in recreational cocaine use rs. J Forensic Sci
1985;30 :873.
8. Ross DL. Factors associated with excited delirium deaths
in police custody. Mod Pathol 1998; II: I I 27.
9. Petty CS, McDonough ET. Positional asphyxia-sudden
death. National Law Enforcement Technology Center Bulletin. U.S. Department of Justice, Nationa l Institute of Justice, June 1995.
10 . McCauley D. Gasping fo r breath. Police 1996;July:56.
I I. Brule D. Combative subjects require special care. Police
1996;July:30.
12. Di Maio OJ, Di Maio VJM. Asphyxia. In Forensic pathology. New York: E lsevier, 1989:217-21.
13. Spitz WU. Asphyxia. In Spitz WU, ed. Medicolegal investigation of death, 3rd ed. Springfield, IL: Charles C
Thomas, 1993 :484-5.
14. Rcay DT. Deaths in custody. Clin Lab Med 1998: I.
15. Mirchandani HG, Rorke LB, Sekula-Perlman A, Hood lC.
Cocaine-induced agitated delirium, forceful struggle, and
minor head injury. Am J Forensic Med Pathol 1994; 15:
95.
16. Karch SB. Cocaine. In Karch SB, ed. ThP pathology of
drug abuse. 2nd ed. Boca Raton, FL: CRC Press, 1996:
1- 175.
17. Karch SB, S tevens B, Ho CI-1. R elating cocaine blood concentrations to toxicity- an autopsy study of 99 cases. J
Forensic Sci 1998;43:41.
J 8 . Karch SB. Amphetamine and methamphetamine. In Karch
SB, ed. Th e pathology ofdrug abuse, 2nd ed. Boca Raton,
FL: CRC Press, 1996: 199-230.
19 . Ruttenber AJ, Lawler-Heavner J, Yin M, Wctli CV, Hearn
WL, Mash DC. Fatal excited delirium following cocaine
use: epidem iologic fin dings provide new evidence fo r
mechanisms of cocaine toxicity. J Forensic Sci 1997;42:
25.
20. Mash DC. Coca ine delirium. In Karch SB, ed. Dmg abuse
handbook. Boca Raton, FL: CRC Press, 1998:414-20.
21. Kosten T R, Kleber HD. Sudden death in cocaine abusers:
re latio n to neuroleptic malignan t syndrome (letter] . Lancet
1987;1: 11 98.
22. Kosten T R, Kleber HD. Rapid death during cocaine abuse:
a variant of the neuroleptic mal ignant syndrome? Am J
Dntg Alcohol Abuse 1988;14:335.
23. Laposata EA, Hal e P, Poklis A. Evaluation of sudden
death in psychiatric patients with special reference to pheno thiazine therapy: forensic pathology. J Forensic Sci
1988;33:432.
24. Cardoff SN, M ann SC. Neuroleptic malignant syndrome.
Mcd Clin North Am 1993;77:185.
25. American Civil L ibenies Union of Southern California.
Pepper spray llpdate: more fatalities. more questions. L os
Angeles: ACLU, June 1995.
26. Granfield J, 01men J, Petty CS. Pepper spray and in-custody deaths. Alexandria, VA : International Association of
Chiefs of Police, March 1994.
27. Ste ffee CH, Lantz PE, Flannagan LM , T hompson RL, Jason DR. Oleoresin caps icum (pepper) spray and "in-custody deaths." Am J Forensic M ed Pathol 1995; 16:183.
28. Ordog GJ, Wasserberg J, Balasubramanium S. E lec tronic
gun (taser) injuries. Ann Em erg Med 1987; 16:73.
Am J Forensic ,..,fed Pmhol. Vol. 21. No. I, March 2000

52

R. L. 0 'HALLORAN AND J G. FRANK

29. Kornblum RN, Reddy S~. Effects of the taser in fatalities
involving police confrontation. J F01;ensic Sci 199 1;36:
434.
30. Ikeda N, Harada A, Suzuki T. H omicidal manual strangulation and multiple stun-gun injuries. Am J Forensic
Med Patholl992;13:320.
3 1. C hute DJ, Smialek JE. Injury patterns in a plastic (AR-1)
baton fata lity. Am J Forensic Med Pat hoi 1998; 19 :226.
32. Bray GA. Obesity: definition, diagnoses a nd d isadvantages. Med J Aust 1985; 142:S2-S8.
33. Raven KP, Reay DT, Harruff RC. A11ifactual injuries of
the larynx produced by resuscitative intubation . Am J Fo rensic Med P athol 1999;20:31.

Am J Forensic Med Pa1hol, Vol. 21, No. 1. March 2000

34. Spitz WU. Asphyxia. In Spitz WU, ed. Medicolegal investigation of death, 3rd ed. Springfield, IL: Charles C
Thomas, 1993:445- 85.
•
35. Bell MD, Rao VJ , Wetli CV, Rodriguez RN. Positional
asphyxiation in adults. Am J Forensic Mrd ?athol I 992;
13:10 1.
36. Decision by Court. Price et al. v. County of San Diego et
al. United States District Court, Sou them District of California . San Diego, 1998.
37. Hirsch CS. Restraint asphyxiation (lette r] . Am J Forensic
Med Pathol 1994;15:266.
38. O'Halloran RL. Authors' response [letter]. IJ.m J Forensic
Med Pathol 1994; 15:348.