Restraint-related cardiac arrest: Pathogenesis, strategies for prevention, and management for hospital clinicians

INTRODUCTION — Physical restraint may be required in the out-of-hospital or medical setting when individuals are perceived to present immediate threat of risk to themselves or others. Individuals have suffered cardiac arrest during the restraint process. The cause is apparent (eg, trauma) in some cases, but a subset never have a clear etiology identified immediately or at autopsy [1]. Deaths of individuals in custody have created controversy surrounding the cause of, and terminology used in, these events [2].

Hospital clinicians, emergency medical personnel, and law enforcement should employ strategies to minimize the risk of cardiac arrest in an agitated patient who requires restraint. Additionally, there are unique considerations to the resuscitation of a patient who suffered a cardiac arrest during the process of restraint or immediately following restraint.

The purported pathophysiology, strategies for prevention, and specific issues in the treatment of a patient with unexplained restraint-related cardiac arrest (RRCA) will be discussed here. The management of the acutely agitated or violent adult, injuries from weapons or asphyxia used during restraint, and general issues in cardiac arrest management are discussed separately.

●(See "Assessment and emergency management of the acutely agitated or violent adult".)

●(See "Evaluation and management of injuries from conductive energy devices (eg, TASERs)".)

●(See "Advanced cardiac life support (ACLS) in adults".)

●(See "Overview of sudden cardiac arrest and sudden cardiac death".)

●(See "Initial assessment and management of the adult post-cardiac arrest patient".)

●(See "Initial management of trauma in adults".)

TERMINOLOGY

●Restraint-related cardiac arrest (RRCA) – Cardiac arrest that occurs during the process of physical control and restraint of an individual that does not have an obvious cause (eg, trauma, strangulation) or explanatory autopsy findings.

●Compression asphyxia – Compression from an external force on the chest and/or abdomen that physically impairs ventilation.

●Restraint-related asphyxia – Decreased ventilation and oxygenation during restraint, usually due to airway obstruction, but compression asphyxia can also contribute.

●Prone restraint – Restraint of the individual lying flat on their chest and abdomen.

●Prone maximal restraint position (ie, hogtie restraint) – A restraint position where an individual is lying prone with handcuffed hands tied to bound feet behind the individual's back.

●"Excited delirium syndrome" or "excited delirium" – These terms have historically been used by death investigators (medical examiners and coroners) as a cause of death when there are no definitive autopsy findings in decedents with delirium and/or agitation immediately preceding the collapse [2,3]. They have also been used in literature on frequency and prevention of death during restraint and safety of certain law enforcement practices. These terms have become increasingly controversial because they potentially exonerate fatal restraint practices. We agree with many professional societies, including the National Association of Medical Examiners, that these terms should not be used [4].

●"Agitated delirium" – A term that has historically been used to describe a patient with agitation and alteration in mental status, often caused by acute intoxication, associated with excitatory physical findings such as hyperthermia or tachycardia (table 1). This is an imprecise term with decreasing acceptance by clinicians. It has been used by forensic specialists and law enforcement as a synonym for "excited delirium," but is more commonly used in the clinical medicine literature [5]. Clinicians may use this term to refer to delirium with hyperactive features (ie, hyperactive delirium), anticholinergic toxicity with delirium, intoxication delirium, drug withdrawal delirium, and other processes.

●Psychosis – Thinking that has lost touch with reality, typically including delusions, hallucinations, and disordered cognition. An individual with untreated psychosis is more likely to act in an agitated and violent manner compared with the general population. (See "Psychosis in adults: Epidemiology, clinical manifestations, and diagnostic evaluation" and "First episode psychosis", section on 'Agitation/aggression'.)

●Agitation – A condition of restlessness, irritability, nervousness, excessive motor activity, hostility, uncooperativeness, and/or excitement that can have multiple causes, including intoxication, withdrawal, psychosis, and many others (table 2). (See "Assessment and emergency management of the acutely agitated or violent adult".)

EPIDEMIOLOGY — Restraint-related cardiac arrest (RRCA) occurs almost exclusively in individuals restrained by law enforcement rather than in patients restrained in the hospital. RRCA is a rare event among the large number of police-citizen encounters that occur each day. Although the incidence is unknown, studies have reported estimates of two to three unexplained restraint-related deaths of individuals in law enforcement custody per year in several populations exceeding 10 million (eg, Los Angeles, Ontario) [6-10].

The phenomena of an agitated individual (commonly due to stimulant use and/or psychiatric illness) who collapses suddenly after a prolonged struggle has been described in psychiatric patients since the 1800s [11,12]. With the widespread misuse of cocaine in the 1980s, the recognized incidence increased significantly, and the term "excited delirium" was first used to describe cocaine toxicity [13,14]. Prolonged physical resistance while restrained is associated with increased mortality; individuals with agitation or agitated delirium rarely experience cardiac arrest in the absence of aggressive restraint [15,16].

An accurate point estimate for the case-fatality rate of agitated patients who require physical restraints does not exist. The nonspecific use of the term "excited delirium syndrome" in the context of describing deaths in custody and the subsequent challenges of defining the at-risk population in studies have led to unreasonably high fatality estimates (eg, greater than 8 percent) based on limited data [6,17,18]. Cases of RRCA are frequently assigned "excited delirium syndrome" as a cause of death by medical examiners, even though no clear diagnostic tools exist, there is no accepted pathophysiologic mechanism, and patients with similar signs and symptoms are routinely managed by emergency medical personnel and emergency clinicians with very low mortality rates [19].

PATHOPHYSIOLOGY

Metabolic acidosis — The most likely pathophysiology of restraint-related cardiac arrest (RRCA) is cardiovascular collapse from severe metabolic acidosis due to isometric muscle contraction, increased catecholamine activity, and inadequate ventilatory compensation from the restraining process [3,20]. However, the underlying pathophysiology remains controversial because of the relative rarity of RRCA and lack of perimortem medical, laboratory, or radiographic evaluations since the individual often dies outside the hospital.

In case series, patients with RRCA who survived to hospital arrival had a severe metabolic acidosis with pH between 6.25 and 7.015 [3,21]. In other case series, the initial cardiac rhythm was bradycardia, asystole, or pulseless electrical activity (PEA) and not a ventricular dysrhythmia, suggesting a metabolic (as opposed to cardiogenic) etiology of the dysrhythmia [6,20]. However, it is difficult to make strong conclusions given the small number of cases, the fact that acidosis occurs commonly in patients who suffer cardiac arrest regardless of mechanism, and that the cardiac rhythm is not immediately available in most patients who suffer RRCA.

Muscles alter metabolism to anaerobic glycolysis during conditions of high-intensity, sustained, isometric activity. Acidosis generation by muscle metabolism and the subsequent compensatory increase in ventilation occur very quickly when restrained. In a volunteer study, simulated physical resistance by striking a heavy bag for 45 seconds caused a significant acidosis that lasted for 10 minutes [22]. Muscle fatigue then results from acidification of the muscle cells caused by accumulation of lactate, inorganic and monovalent phosphate, and adenosine diphosphate. Maximal acute exercise to exhaustion is associated with systemic pH <7 and serum lactate concentrations >20 meq/L. (See "Energy metabolism in muscle", section on 'Anaerobic glycolysis'.)

Agitation and intoxication — Intoxication with a sympathomimetic agent (eg, cocaine, amphetamine) and/or psychomotor agitation may predispose to RRCA during a law enforcement encounter. Restrained patients who are intoxicated with a sympathomimetic agent appear to be at particularly high risk for adverse outcomes.

Patients who are agitated due to intoxication or delirium are less likely to stop fighting against restraints in response to the pain signaling created by acidosis or injury [20]. In these circumstances, the restrainer commonly uses prolonged and more forceful restraint techniques and may perceive a need for prone positioning.

Sympathomimetic intoxication can also predispose to the development of and exacerbate a metabolic acidosis. Sympathomimetic agents increase catecholamine activity and stimulate catabolic processes that generate ketones and lactic acid. Sympathomimetic-induced psychomotor agitation contributes to muscle generation of lactic acid. Increased sympathetic tone causes vasoconstriction and impedes clearance of metabolic waste products. (See "Cocaine: Acute intoxication" and "Methamphetamine: Acute intoxication".)

Evidence supports that sympathomimetic intoxication is a strong risk factor for RRCA. In case series, most patients who suffered unexplained cardiac arrest during restraint had either acute or chronic misuse of a sympathomimetic agent [6,21]. The incidence of RRCA began increasing in the 1980s, which coincides with the increase in cocaine misuse in the United States [23]. Several case series that included post-mortem toxicology analysis have described high rates of sympathomimetic agents identified in decedents with RRCA [24-27].

Role of positioning — Positioning and weight applied to the torso during restraint can cause limitation of ventilation that likely contributes to RRCA. However, positioning and torso constriction do not appear to cause hypercapnia in healthy individuals and are not solely responsible for RRCA. Altered respiratory mechanics due to a prone posture restrict maximal respiratory compensation in a patient with acidosis, constraining the ability to correct the acidosis. The early association that was described between prone positioning and fatal RRCA has been reported even in cases where signs and symptoms of "excited delirium" are not present [7,10,26].

Studies on healthy volunteers in controlled environments have shown limited but measurable physiologic effects of prone positioning. The prone position itself decreases ventilation 8 to 16 percent, while the prone maximal restraint position (ie, hogtie) decreases ventilation by 13 to 40 percent [28-32]. This position-related effect occurs from mechanical restriction of chest wall movement, unlike the physiologic effects for hypoxemia due to coronavirus disease 2019 (COVID-19) or acute respiratory distress syndrome, in which prone positioning makes ventilation more homogeneous, reduces alveolar distension, and recruits alveoli that had collapsed during supine ventilation. Applying weights to the back further decreases maximal ventilation by approximately 12 to 21 percent [28,29,33]. Although such physiologic effects do not cause hypoxia or hypercapnia in healthy subjects, the limitations induced by these measurable physiologic changes may become critical in a patient with extremely increased metabolic and catecholaminergic demands from agitation or intoxication.

Prone positioning can also impair pulmonary clearance of carbon dioxide by decreasing cardiac venous return, thus decreasing pulmonary perfusion. Prone positioning causes increased pressure on the inferior vena cava, right atrium, and other mediastinal structures. Decreased cardiac output is a predictable physiologic effect, as has been demonstrated in patients undergoing anesthesia in the prone position [34]. While adverse outcomes from such hemodynamic effects in healthy individuals are rare, they may be significant in a patient with extremely increased metabolic demands and concomitant ventilation impairment. (See "Patient positioning for surgery and anesthesia in adults", section on 'Physiologic effects of prone positioning'.)

Role of conductive energy device discharge — Conductive energy devices (CEDs) such as the TASER have been temporally associated with sudden, unexplained deaths of individuals in custody [35]. The direct cardiac effects of the CED electrical discharge are unlikely to cause cardiac arrest, but in certain circumstances, the discharge could be hypothesized to exacerbate a dysrhythmogenic metabolic acidosis. Volunteer studies have shown that CED exposures lasting less than 15 seconds increase lactate and catecholamines and decreases blood pH and minute ventilation; these are likely clinically insignificant physiologic changes under normal conditions [36,37]. However, these mild alterations may be less manageable in a patient with a severe metabolic acidosis and limitations of ventilatory compensation. (See "Evaluation and management of injuries from conductive energy devices (eg, TASERs)".)

Other theories — Despite the evidence supporting the above mechanism, experts dispute the contribution of prone positioning by referring to the overall low incidence of adverse cardiac events of individuals in prone restraint [38]. The underlying pathophysiology of RRCA remains controversial [39]. The following alternative theories have been proposed for why some individuals, based on acquired or pre-existing conditions, may be sensitive to restraint:

●Death may occur solely from the catecholamine surge caused by stimulant intoxication. However, in postmortem studies, unexplained fatalities from preterminal "excited delirium syndrome" were found to have serum cocaine concentrations 3- to 10-fold lower that those typically associated with death from cocaine toxicity [7,14,40,41]. Thus, while a catecholamine surge could be contributory in some cases, it is not likely to be directly causative in an otherwise healthy, unrestrained individual.

●A stress cardiomyopathy (ie, Takotsubo) from prolonged sympathetic hyperactivity may contribute or be causative [42]. However, unlike cases of RRCA, Takotsubo typically affects older females and is rarely lethal. (See "Clinical manifestations and diagnosis of stress (takotsubo) cardiomyopathy".)

●A pre-existing cardiomyopathy may be contributory. For example, cardiomyopathy (hypertrophic or dilated) is often found on autopsy of RRCA patients. Postmortem studies of RRCA cases found a cardiomyopathy in approximately one-third of decedents [6,27]. Likewise, dilated cardiomyopathy has been reported among individuals who misuse cocaine [43,44]. A cardiomyopathy, if present, is likely a contributing factor rather than causing or developing from the RRCA. Left ventricular hypertrophy or a dilated cardiomyopathy can predispose patients to a dysrhythmia, especially in the presence of increased catecholamines, acidosis, and/or other metabolic derangements. (See "Clinical manifestations, diagnosis, and management of the cardiovascular complications of cocaine abuse", section on 'Cardiomyopathy'.)

●Victims may have a variant form of neuroleptic malignant syndrome (NMS) due to central nervous system dysfunction of dopamine signaling, possibly caused or exacerbated by cocaine or other stimulant misuse [12]. NMS is typically precipitated by first-generation antipsychotic agents, but an NMS-like syndrome is associated with other etiologies, such as abrupt cessation of dopamine agonists (eg, levodopa, amantadine) [45,46]. Cocaine inhibits presynaptic reuptake of dopamine. Chronic cocaine misuse is associated with neurobiological adaptations such as upregulation of dopamine transporters and function [12]. Other sympathomimetic agents (eg, methamphetamine, cathinones) that also inhibit dopamine reuptake have been associated with RRCA [40,47]. A failure of dopamine transporter upregulation is postulated to predispose to RRCA by causing excess synaptic dopamine and autonomic dysfunction (ie, catecholamine surge leading to a hypermetabolic state, hypertension, vasospasm, and cardiac arrhythmogenicity) [12]. A postmortem study found that patients who died from cocaine-related RRCA had fewer presynaptic dopamine transporters compared with age-matched controls with chronic cocaine misuse [40].

●Victims who chronically misuse cocaine may have increased central nervous system oxidative stress from NADPH oxidase 2 enzyme dysregulation [48]. Although the molecular mechanisms are not completely understood, these individuals are postulated to be predisposed to cocaine-induced neurotoxicity, which may lead to alteration of several systems such as heat shock protein generation and inhibitory (eg, gaba-aminobutyric acid) neurotransmission [40,48].

CLINICAL PRESENTATION — Individuals who suffer from restraint-related cardiac arrest (RRCA) are overwhelmingly young, are male, have a high body mass index, and have a history of substance abuse and/or psychiatric disease. Most are restrained prior to hospital arrival by law enforcement or emergency medical personnel, but agitated, restrained hospital patients are also at risk for RRCA.

RRCA is strongly associated with prone positioning, especially with pressure applied to the back. In a case-control study of restrained patients, those who had cardiac arrest were more likely to be positioned prone (95 versus 67 percent) and experience thoracic compression (75 versus 24 percent), although tests of statistical significance were not presented [9]. Pressure and positioning may play a role in otherwise unexplained deaths even when not severe enough to cause compression asphyxia and death by themselves. (See 'Role of positioning' above.)

Prior to cardiac arrest, law enforcement has typically described individuals as displaying some combination of the following signs and symptoms: agitation, aggressive or violent behavior, inappropriate removal of clothing, attraction to shiny or glass objects, diaphoresis, hyperthermia, and tachypnea [49,50]. They often also have signs of sympathomimetic intoxication (table 1). (See 'Agitation and intoxication' above.)

Individuals who suffer RRCA are frequently described as struggling against restraint followed by sudden unresponsiveness and cardiopulmonary arrest [3]. The initial rhythms documented by emergency medical personnel are usually asystole or pulseless electrical activity (PEA), although data are limited.

STRATEGIES FOR PREVENTION

Avoid escalation, if possible — The best way to avoid restraint-related cardiac arrest (RRCA) is to avoid physical restraint in the first place, if possible. Verbal de-escalation and/or rapid administration of calming medications to control agitated behavior can limit the need for physical restraints. (See "Assessment and emergency management of the acutely agitated or violent adult", section on 'Verbal techniques'.)

If physical restraints are likely to be needed, anticipation prior to escalation can lead to limited use of force and risk of injury or death. (See "Assessment and emergency management of the acutely agitated or violent adult", section on 'How to apply restraints'.)

Appropriate use of restraint — Once all attempts have been made to avoid physical restraint, the agitated patient must be restrained as safely as possible. The purpose of restraint is to protect the patient and treatment team and to facilitate a medical evaluation. Restraints should be removed as soon this can be safely accomplished.

●The restraint team should have at least five people to each control an extremity and the head.

●We recommend avoiding prone restraint and rapidly turning a patient supine if prone positioning occurs during the restraint process.

●We suggest rapidly evaluating a patient brought to the hospital restrained in a prone position by emergency medical services or law enforcement. The patient should be moved out of the prone position as soon as possible since RRCA can occur quickly, especially if the patient is struggling against the restraints. This is typically accomplished by administering calming medications; in rare circumstances, rapid sequence intubation and tracheal intubation may be needed if the patient cannot be sedated and continues to struggle against restraints.

Brief deviations from these practices may be necessary under certain circumstances to protect the patient, those providing restraint, or others nearby. Restraint of an agitated patient is a dynamic process that requires constant evaluation and balancing of risks to the patient, staff, first responders, and anyone in the patient's immediate surroundings (eg, other patients). (See "Assessment and emergency management of the acutely agitated or violent adult", section on 'Physical restraints'.)

Administer medications to control agitation — Rapidly administer calming medications to control agitated behavior and facilitate decreased struggle against physical restraint. A treatment algorithm outlining an approach to selecting medications is provided (algorithm 1), and evidence regarding the efficacy and safety of various medications is discussed separately. (See "Assessment and emergency management of the acutely agitated or violent adult".)

Avoid any pressure on neck — We recommend not applying any pressure on the neck or using "choke holds" during the restraining process. Even partial airway obstruction can cause hypoxemia or impair the respiratory compensation of the metabolic acidosis associated with RRCA. Vascular neck restraints and "choke holds" have been used controversially by law enforcement as a means to obtain rapid unconsciousness and/or compliance during restraint, but such tactics have no role in the medical setting [51,52].

Recognize "air hunger" as a potential warning sign — A restrained patient who says "I can't breathe" may be at risk of imminent cardiovascular collapse, and immediate corrective measures should be initiated, such as the following:

●Measure oxygen saturation with pulse oximetry. It is important to note that a normal oxygen saturation can be falsely reassuring. If the patient has a metabolic acidosis with inadequate respiratory compensation, the sensation of dyspnea may be an indication of hypercapnia and acidosis rather than hypoxia. (See "Physiology of dyspnea", section on 'Acute hypercapnia'.)

●Rapidly turn a patient supine if prone, removing any pressure on the neck, chest, or abdomen; ensure there are no signs of airway obstruction; and administer medications to control agitation, if necessary. (See 'Strategies for prevention' above.)

In a case report, a patient said "I can’t breathe" shortly before RRCA [3]. Media reports of numerous high-profile RRCA cases describe a similar association. However, restrained patients are often reported to say "I can't breathe" without suffering RRCA, and there is likely a multimodal etiology of the dyspnea sensation. The true incidence of hypercapnia in restrained patients is unknown and likely not the cause for dyspnea in most cases.

Close monitoring after patient calm — In a patient who was restrained prone or was struggling against restraints for more than five minutes, we begin continuous cardiac monitoring for at least one hour, frequently re-evaluate clinical status, exclude hyperthermia, and perform a full evaluation for signs of trauma once it is safe to do so. (See "Assessment and emergency management of the acutely agitated or violent adult", section on 'Monitoring and care of physically restrained patients'.)

If the cause of prior agitation is not apparent (eg, intoxication, withdrawal, psychiatric), further investigation is required to exclude an etiology that may require medical or surgical intervention. (See "Assessment and emergency management of the acutely agitated or violent adult", section on 'Post-restraint medical evaluation'.)

MANAGEMENT OF CARDIAC ARREST

General management — The general management of a patient in restraint-related cardiac arrest (RRCA) is similar to that for other causes of cardiac arrest and discussed separately. (See "Adult basic life support (BLS) for health care providers" and "Advanced cardiac life support (ACLS) in adults" and "Pediatric basic life support (BLS) for health care providers" and "Pediatric advanced life support (PALS)".)

In a patient with pulseless electrical activity (PEA) or asystole due to RRCA, we suggest administering sodium bicarbonate in addition to other standard therapy (eg, epinephrine). We administer intravenous (IV) sodium bicarbonate 7.5 or 8.4% 50 to 100 mL over one to two minutes (ie, one to two ampule pushes) as part of the resuscitation. This dose can be repeated every three to five minutes concurrently with epinephrine administration. (See "Advanced cardiac life support (ACLS) in adults", section on 'Asystole and pulseless electrical activity'.)

Despite recommendations against the routine use of sodium bicarbonate for cardiac arrest, selective use of sodium bicarbonate is reasonable when there is clinical suspicion or laboratory evidence of significant pre-existing metabolic acidosis, which is very likely in a patient with RRCA. Supporting evidence is based on case reports of patients surviving RRCA with IV bicarbonate, the author's clinical experience, and use of bicarbonate for resuscitation in other diseases that cause a severe metabolic acidosis (eg, ethylene glycol poisoning) [21,53]. (See "Bicarbonate therapy in lactic acidosis" and "Advanced cardiac life support (ACLS) in adults", section on 'Other medications'.)

Post-arrest care — In a patient with return of spontaneous circulation after RRCA, management should focus on stabilization and preventing rearrest. The general management of a post-cardiac arrest patient is discussed separately. (See "Initial assessment and management of the adult post-cardiac arrest patient".)

Until the results of the initial post-arrest blood gas are available, we assume the patient has a metabolic acidosis. If the patient requires tracheal intubation, we avoid neuromuscular blockade, if possible, so they can initiate breaths on the ventilator and meet their minute ventilation requirements. We ensure the patient has adequate ventilation guided by blood gas results. In spontaneously breathing patients with severe acidemia, an appropriate ventilatory response to the metabolic acidosis should reduce the partial pressure of carbon dioxide (PCO₂) to ≤15 mmHg. (See "Overview of initiating invasive mechanical ventilation in adults in the intensive care unit", section on 'Settings'.)

In a patient with a severe metabolic acidosis (pH <7.1) after adequate ventilation is ensured, we administer intermittent IV boluses of sodium bicarbonate (1 to 2 mEq/kg body weight). A suggested approach is discussed in detail separately. (See "Bicarbonate therapy in lactic acidosis", section on 'Approach'.)

Measuring the patient's core temperature is required, followed by cooling measures as needed. A patient may feel "warm to the touch" from vasodilation, but hyperthermia with core body temperatures ≥40.5°C (105°F) is a common finding [40]. Cooling measures for hyperthermia are discussed separately. (See "Severe nonexertional hyperthermia (classic heat stroke) in adults", section on 'Management'.)

LEGAL CONSIDERATIONS — The following are suggestions for a clinician who manages a patient who suffers a restraint-related cardiac arrest (RRCA) while in law enforcement custody:

●Document who provided the history, the circumstances of the RRCA, and a full trauma examination. Do not speculate in the documentation. Expect that you may be called to testify.

●Be aware of any local regulations on mandatory reporting of fatalities that occur while in restraints.

●If there are any uncertainties, reference hospital and local policies and contact the hospital's legal counsel.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Adult with altered mental status in the emergency department".)

SUMMARY AND RECOMMENDATIONS

●Terminology – Restraint-related cardiac arrest (RRCA) is cardiac arrest that occurs during the process of physical control and restraint of an individual that does not have an otherwise obvious cause (eg, trauma, strangulation) or explanatory autopsy findings. (See 'Terminology' above.)

●Epidemiology – RRCA is rare and occurs more frequently in individuals restrained by law enforcement than in patients who are restrained in the hospital. Prolonged physical resistance while restrained increases mortality; individuals with agitation or agitated delirium rarely experience cardiac arrest in the absence of aggressive restraint. (See 'Epidemiology' above.)

●Pathophysiology – The most likely pathophysiology of RRCA is cardiovascular collapse from severe metabolic acidosis due to isometric muscle contraction, increased catecholamine activity, and inadequate ventilatory compensation. There is a strong association with stimulant intoxication and prone positioning. (See 'Pathophysiology' above.)

●Clinical presentation – Individuals who suffer RRCA are overwhelmingly young, are male, have a high body mass index, and have a history of substance abuse and/or psychiatric disease. They are frequently described as struggling against restraint followed by sudden unresponsiveness and cardiopulmonary arrest, with an initial rhythm of asystole or pulseless electrical activity (PEA). (See 'Clinical presentation' above.)

●Strategies for prevention

•Avoid physical restraint, if possible, with verbal de-escalation and/or rapid administration of calming medications. A treatment algorithm outlining an approach to selecting a medication to control agitated behavior is provided (algorithm 1). (See 'Avoid escalation, if possible' above and 'Administer medications to control agitation' above.)

•Avoid prone restraint and rapidly turn a patient supine if prone positioning occurs during the restraint process. (See 'Appropriate use of restraint' above.)

•Do not apply any pressure on the neck or use a "choke hold" during the restraining process. (See 'Avoid any pressure on neck' above.)

•A restrained patient who says "I can't breathe" may be at risk of imminent cardiovascular collapse, and immediate corrective measures should be initiated. (See 'Recognize "air hunger" as a potential warning sign' above.)

•For a patient who was restrained prone or was struggling against restraints for more than five minutes, we place them on continuous cardiac monitoring for at least one hour, frequently re-evaluate clinical status, exclude hyperthermia, and perform a full trauma evaluation once it is safe to do so. (See 'Close monitoring after patient calm' above.)

●Management of patient in cardiac arrest – The general management of a patient in RRCA is similar to that for other causes of cardiac arrest and discussed separately. (See "Adult basic life support (BLS) for health care providers" and "Advanced cardiac life support (ACLS) in adults".)

In a patient with PEA or asystole due to RRCA, we suggest administering sodium bicarbonate, in addition to other standard therapy (ie, epinephrine) (Grade 2C). We administer intravenous (IV) sodium bicarbonate 7.5 or 8.4% 50 to 100 mL over one to two minutes (ie, one to two ampule pushes) as part of the resuscitation. This dose can be repeated every three to five minutes concurrently with epinephrine administration. (See 'Management of cardiac arrest' above.)

●Post-cardiac arrest care – In a patient with return of spontaneous circulation, management should focus on stabilization and preventing rearrest. If the patient is tracheally intubated, we avoid neuromuscular blockade (if possible) to preserve respiratory drive. Measuring the patient's core temperature is required, followed by cooling measures as needed. (See 'Post-arrest care' above.)

In a patient with a severe metabolic acidosis (pH <7.1) after adequate ventilation is ensured, we administer intermittent sodium bicarbonate boluses (1 to 2 mEq/kg IV). The evidence is discussed separately. (See "Bicarbonate therapy in lactic acidosis", section on 'Approach'.)