Restraint-Related Positional Asphyxia “Near-Death”
A Case Study

This Case Study was originally posted in August of 1998, days after the event occurred.
In December of 2006, I began rewriting this Case Study, so as to make a “PRINTER FRIENDLY” pdf file of it.
And, on January 1st, 2007, I POSTED its NEW VERSION!:

Restraint Asphyxia Near-Death Case Study

The “original” version of this Case Study remains posted on the pages it always has been.
(Below and on 6 other pages.) But, the NEW VERSION is ever so much BETTER!
And, it’s certainly better for PRINTING purposes.

(OLD) Restraint-Related Positional Asphyxia “Near-Death”
A Case Study

This Patient Entered Respiratory Arrest Following
Forceful Restraint in a Prone Position for Approximately 3 Minutes,
And Hobble Restraint For Less Than 1 Minute!


Denver Health Medical Center (formerly known as “Denver General”) Paramedic Division is the sole provider of 911 EMS and advanced life support for the City and County of Denver, Colorado. Each 911 ambulance is staffed with two paramedics. We employ a “Dynamic Dispersal” posting system, wherein our ambulances are perpetually on the streets, posted at geographically strategic locations when not running a call. We do not operate from stationary quarters, but are mobile, in our vehicles, the entirety of each 10-hour shift.


On a day in August, 1998, soon after 0100 hours, my partner and I were posted in our ambulance, parked at a convenience store. [A “7-11!”] We happened to be monitoring the police channel for our posted district (something not required and not always done either by myself or others). Some time around 0100 hours, I heard the police dispatched to investigate the report of "a man running around naked." Because of my acquaintance with situations involving restraint-related positional asphyxia, I immediately turned up the radio volume. I knew that such a situation is often the precursor to restraint asphyxia.

At approximately 0132 hours, we heard a police officer radio for an ambulance, and a requesting a Denver Police Department (DPD) supervisor response to an address approximately two blocks away from the original investigation location. The officer was panting and breathing very hard, as though he’d just completed a strenuous chase or struggle. He requested the ambulance to respond "code nine" (nonemergent), indicating that the patient’s medical need appeared (to him) to be minor. At that time, at my direction, my partner and I “self-dispatched” ourselves, going enroute to the scene, without employing lights and sirens. The distance between our posted location and the scene was 10 city blocks.

At 0133, DPD dispatchers logged the police officer’s call for an ambulance. (It is likely that they logged the call for the supervisor first, thus the time lag between the officer’s call and logging the ambulance request.)

At 0133.15 Ambulance dispatchers received the call information from DPD dispatch.

At 0134.11 As we were about a block away from the scene, we began to receive dispatch information from our dispatchers.

As we pulled up to the scene (while copying the dispatch information), I observed several police officers restraining a naked person, face-down, in the grass of a residential front lawn. Two or three officers were holding the patient’s shoulders, trunk and hips, lower legs down. Another officer was clearly observed to be kneeling, with his left knee pressing down upon the patient’s head – the patient’s face was turned away from us. Yet another officer was just completing attachment of the hobble restraint strap, connecting the patient’s handcuffed wrists to his hobble-strap-tied ankles, as we pulled up to the scene.

At 0134.43 As I acknowledged receipt of our dispatch information, I immediately logged our arrival “on scene,” and stepped out of the ambulance.

As I opened my door, I noticed that there was no noise (yelling or otherwise) coming from the patient. I immediately directed the officers to “Hey! Roll him to his side, guys!” The officers’ approximate reply (while rolling the patient to his left side) was, "OH! He’s OK. … He’s breathing! … He was just yelling a second ago!"

Glancing at the now-exposed anterior of this young, Caucasian, male individual, I noticed presence of minor facial bleeding, the lack of chest or abdominal trauma, that his torso was a mottled pale/purple color, and that his face was clearly cyanotic (in night-lighting). I also noticed NO signs of respiratory effort or sounds of breathing. I immediately feared that the patient was already dead.

As I bent to check for a pulse, I directed the officers to, "Get everything off him, NOW, guys!"

Finding a carotid pulse of approximately 120/minute, I realized the patient was still alive, but in respiratory arrest. Quickly observing the patient’s oropharynx, I determined that there were no obvious foreign bodies present, and simply acted to hold his airway open, using a jaw-thrust airway maneuver.

My partner had already brought our wheeled stretcher (pram/cot/gurney) to the patient’s side as the officers completed removal of the hobble restraint ankle tie and the handcuffs. (Our vehicle was only approximately 5 to 10 feet from the patient – hence my decision to move the patient to our ambulance for further treatment.)

We placed the patient supine on our wheeled stretcher. This 18 to 20 year old male was unconscious, flaccid, and in respiratory arrest. His left eye was open with a fixed, unfocused gaze – his right eye was dramatically swollen-shut by soft tissue trauma. I engaged the chest restraint high up on the supine patient’s chest (beneath his arms), and pulled it tight – then resumed the jaw-thrust airway. The lower body restraint strap was engaged just above the patient’s knees at my direction, and pulled very tight.

Just as we were loading the patient/wheeled stretcher into the ambulance, I heard a gasping, gurgling noise and observed a slow, deep, and non-rhythmic rise and fall of the patient’s chest – indicating a spontaneous return of respiratory effort. Some non-purposeful patient extremity movement soon began, and soft bilateral wrist restraints were applied, using double-folded, strong, Kerlix gauze. The patient’s right wrist was restrained and pulled “up” to be attached to the central T-junction at the top of the pram frame. The patient’s left wrist was tractioned straight, and secured to the left side of the wheeled stretcher frame at a T-junction distal to his wrist.

As my partner prepared to obtain IV access and I prepared intubation equipment, I did a quick visual exam of the patient’s anterior body. He had right orbit soft tissue trauma, with both upper and lower right eyelids darkly bruised and swollen tightly shut. There was a small (? sutureable) laceration on his upper right eyelid, and he had a minor (already drying) bilateral epistaxis. His left eye was frequently open, but his gaze was unfocused and he remained unresponsive. His left pupil was dilated and reacted slowly to light. His right eyelids were too swollen for me to part enough to observe his right pupil’s response to light. Thus, I had no indication of whether the patient’s persistent altered level of consciousness might be due to head trauma, due to drug intoxication, due to post-respiratory arrest effects, or due to a combination of all the above.

By this time, the patient’s respiratory efforts had increased to 26 or 30 breaths per minute. Audible noises indicating fluid in the airway were present. I suctioned a small amount of blood and mucous from his oropharynx, immediately correcting these audible airway-fluid noises.

Because of his right orbit trauma, I elected to use his left nare for nasal intubation. I prepared his left nare by instilling neosynephrine spray and lidocaine gel into same. With the patient’s good respiratory effort, and his lack of resistance to (awareness of) the painful stimulation of nasal intubation, the patient was quickly successfully nasally-intubated with a lubricated #7.5 endotracheal tube. Good tube placement was verified by obvious respiratory movement of air via the tube (creation of mist within the tube upon each exhalation), and the absence of audible air movement via the oropharynx. Additionally, the patient’s skin condition was now no longer mottled or cyanotic in areas – but pale, cool, and moist overall.

With his elevated spontaneous respiratory effort and lack of obvious chest trauma, I elected to administer oxygen via a non-rebreather apparatus attached to the endotracheal tube via a "T-piece" connector, and run at 15 LPM. Auscultation of the patient’s lungs revealed clear, equal bilateral breath sounds with full, equal, bilateral chest excursion. While securing the nasotracheal tube, I radioed for Denver Fire Department (DFD) assistance (in the event that the patient developed a need for assisted ventilation.)

Meanwhile, my partner had obtained IV access with a 14 gauge angiocath in the patient’s left forearm. A full set of blood samples were obtained, and Normal Saline was run at approximately 120 drops/minute via a Blood-Y (trauma) infusion apparatus.

After completion of intubation and in-line oxygen administration, the patient’s skin was noted to be cool, dry, and PINK. Soon after that, the patient began to move his head back and forth – his activity becoming more and more forceful, in an apparent effort to extubate himself. Endotracheal tube placement became threatened by the patient’s combative head activity.

I initiated manual restraint of the patient’s head, by grasping his bilateral jaw angles and applying a minimum of “in-line traction” to prevent head movement, while simultaneously restraining his jaw in a “jaw thrust” position. This maneuver was successful, but only with a great deal of physical effort on my part (the patient’s efforts to thrash his head about were impressively strong, and becoming stronger by the minute).

My partner administered Narcan, 2mg, IV without change in the patient’s combativeness. Next, he administered Dextrose 50% 25 G. IV without change in the patient’s combativeness.

By this time, DFD arrived and I directed a firefighter to assume manual jaw/head control in the same manner that I had maintained it. Re-auscultation of the patient’s chest revealed continued good BS bilaterally (appropriately accompanying corresponding NRB bag deflation/inflation) with continued misting of the ET tube on exhalation – re-confirming good ET tube placement.

We went enroute to the ED at 0151.57 hours, using lights and sirens. The patient’s blood pressure was 174 systolic, by palpation. I notified the receiving hospital’s emergency department (ED) of our patient’s condition and our estimated time of arrival.

During transport, the patient’s ECG showed a sinus rhythm at a rate fluctuating between 150 and 160/minute. His electrical axis was normal, his QRS complexes and S-T segments were within normal limits, and no ectopy was observed.

Once enroute to the ED, I adjusted the IV flow rate down to TKO. I employed several verbal cue directions and determined that the patient was “awake,” but remained unresponsive to verbal stimuli (unable/unwilling to follow cues). The patient persisted in very forceful attempts to combatively defeat extremity restraints and manual jaw/head restraint enroute. The placement of each wheeled stretcher safety-belt-restraint and the bilateral soft wrist restraints were not defeated.

The persistent flexing and relaxing of the patient’s arms throughout transport interfered with obtaining an auscultated blood pressure.

Enroute to the ED, the persistent forcefulness of the patient’s attempts at head movement required the first head-restraining firefighter to switch-off with another head-restraining firefighter, secondary to finger fatigue (pain)! In preparation for ED patient care transfer, I restrained the patient’s ankles together using Kerlix gauze.

At 0157.54 hours we arrived at the ED, having traveled a distance of 38 city blocks. There were no further changes in the patient’s activity, orientation (cue-following), or medical condition upon arrival.

[NOTE: This patient had facial trauma and an altered level of consciousness, thus suggesting the possibility of spinal injury mechanism. In hind-sight, I obviously should have applied a cervical collar and attempted taped-head- and full-body-immobilization of this patient to a long back board or scoop stretcher. However, in spite of my lack of C-collar application and taped-head-immobilization, my care of this patient consisted of full-body restraint (immobilization) to my wheeled stretcher, with continued manual restraint (immobilization) of the patient’s head. It is my opinion that, the amount of combative head-thrashing exercised by this patient would NOT have been controlled by a C-collar and taped-head-immobilization, anyway! Thus, my care provision barely satisfied the basic requirements of spinal immobilization precautions … And, I know better NOW!]


The above time factors support the following conclusions:

Published studies of police-involved deaths attributed to restraint-induced positional asphyxia do not consistently document the time lapse between placement in restraints and onset of noticed unconsciousness or death. However, of five cases that do report this time period [two cases from Reay et al,(1) and three cases from O'Halloran and Lewman(2)], the average time between restraint application and onset of death was 5.6 minutes! Thus, a time frame of less than 4 minutes between initiation of forceful-prone restraint and respiratory arrest is entirely consistent with previous studies of restraint asphyxia death.


A discussion that reveals the emergency department findings, and explores the pathophysiology of this case’s restraint-related near-asphyxia, is provided on “Page Two” of this case study.


This counter was started on July 1st, 2001.