Perspectives on Excited Delirium and Ketamine

Psychiatric Times, Vol 39, Issue 3,

Parenteral ketamine is commonly used to manage agitation that is deemed severe and obstructive to necessary medical evaluation and treatment. But what does this suggest about the relationship between medicine and society?


Over the past 2 decades in psychiatry, interest in ketamine has grown based upon the potential therapeutic benefit of subanesthetic doses to rapidly improve symptoms of otherwise treatment-resistant depression.

During roughly the same period, use of the dissociative agent has expanded to emergency medicine and related areas of critical care medicine. In prehospital, emergency department, and intensive care settings, it has become more common to see high doses of parenteral ketamine used to manage agitation that is deemed severe and obstructive to necessary medical evaluation and treatment. Emergency medicine practitioners coined the term “excited delirium syndrome” (ExDS) in an American College of Emergency Physicians (ACEP) white paper to identify a subset of patients whose presenting phenotype correlates with the possibility of having an underlying critical illness.1

The authors of this article have lived, learned, and worked in a US urban community in which the expansion of ketamine for such situations has outpaced much of the rest of North America. Based on our combined expertise, we hope to provide useful context for this practice seated in its history with attention to the evolution of nosology and with acknowledgment of changes in the relationship between medicine and society.

Historical Context: Bell's Mania

The entity ExDS is said to have its earliest roots in the condition of Bell's mania: exceedingly great overactivity of sudden onset with manifestations of manic symptoms, delusions, and hallucinations, and the appearance of confusion.2 In an 1849 article in the American Journal of Insanity, Luther Vose Bell, MD, may have reported the latter because patients were so energetic and difficult to engage in dialogue that clinicians might find it challenging to adequately assess attention, orientation, and other features of mental status to ascertain if they were experiencing a delirial state. In light of that conundrum, Bell's mania has also been called delirious mania.

Bell, and others following him, noted that without treatment this condition can be fatal; the recommended treatment was “sufficient sedative,” and it was suggested that even large doses might not yield the sleep patients needed to recover. Since medicine conceptualizes delirium not as a psychiatric disorder but rather as a syndromic reflection of the effects of a severe medical condition on the whole body including the central nervous system, there is face validity to the notion that delirious mania could be deadly. If the underlying condition for ExDS might also be life threatening, it would make sense that provision of potent sedation by modern medical means would be the order of the day.

Although it is clinically informative to understand the historical connection between ExDS and Bell's mania, that connection also highlights aspects that are problematic. Bell indicated that patients in his cohort were able to “suspend the intellectual wanderings long enough to answer correctly as to orientation” if a clinician insisted upon having the mental status queries answered. Although the impairments inherent to delirium may wax and wane, such a description suggests that that historical cohort may have been experiencing psychotic mania, but not delirium. Also, although it was recognized that the condition could be life threatening, his patients died after a 3- to 6-week course of illness, apparently due to cardiovascular collapse secondary to unmitigated overactivity. ExDS is conceptualized as a condition of much more acute danger with an underlying disease process that could result in death rapidly—within minutes to hours of its abrupt onset.1

Furthermore, although patients with ExDS may be delirious, the nomenclature more aptly reflects the difficulty in determining whether such agitated patients have delirium rather than accurately indicating that they do. The criteria for ExDS (Table) do not align with the DSM-5 conception of the syndrome of delirium or any previous version thereof.

The term excited delirium (sometimes used interchangeably with agitated delirium) is a fusion of a descriptive participle used to communicate something about behavioral energy and a term with a different, preestablished meaning in the nomenclature of psychiatry. And since agitation is used commonly and imprecisely across a variety of medical settings, clarity about the nature and utility of ExDS is elusive. Thus, we point out the historical connection between Bell's mania and ExDS to indicate how the most important consequence may not be to demonstrate progress in medicine, but instead to allow imprecise thinking and use of language with the potential to stigmatize.

Intoxicating Substances and the History of ExDS

The connection between the history of psychiatry and current practice around ExDS reveals parallels that can be helpful in understanding the evolution of the emergency medicine concept and debates about not only its proper management but also its very existence. ExDS is, as the term indicates, a syndrome. Similar to the whole collection of psychiatric disorders, ExDS is not a diagnosis with a defined pathophysiology. Like our DSM entities, ExDS is defined by criteria set forth by a group of expert practitioners and taken up by its specialty field for use to describe a recognized problem. ExDS has more direct roots within emergency medicine itself, specifically in the subspecialty realm of medical toxicology.

It has been known for decades that patients experiencing the effects of severe intoxication with stimulants, specifically cocaine, are at acute risk of morbidity or death. Massive monoaminergic activity causes autonomic stress on the cardiovascular system and elevates body temperature. Stimulation of motivated behavior along with accompanying perceptual disturbances may drive excessive physical activity that increases anaerobic metabolism and physiologic stress. Cocaine also acts as a sodium channel antagonist, thereby adding to the risk of seizures and fatal cardiac arrhythmias. The established treatment for patients experiencing these toxic effects involves liberal doses of benzodiazepines to provide calm and protect against aberrant conduction, along with fluids, cooling measures, and supportive critical care.

Medical evaluation and testing of these patients will reveal findings consistent with the established pathophysiology that points to a diagnosis of cocaine intoxication. Note, however, that one cannot put this medical knowledge into targeted practice for a given patient without close and ongoing physical interaction with the individual experiencing the condition. And patients experiencing cocaine intoxication generally do not simply present themselves to emergency departments for care because of chest discomfort or hallucinosis. They may be first encountered in the prehospital setting by other members of the community and, next, by law enforcement professionals.

Although difficult to quantify, teaching police about cocaine undoubtedly saved many individuals from dying in squad cars and/or in jail—especially in cases of “body stuffing” where an alleged perpetrator of illegal drug activity quickly ingests the evidence to avoid criminal charges. Even when paramedics arrive and help convert the focus from law to medicine, such individuals do not automatically become patients in the traditional sense by engaging with emergency medical services personnel for assistance and care. Vital signs, physical findings, electrocardiograms, and laboratory results must be gathered and critically evaluated to render a diagnosis of stimulant toxicity. As a result, there must be a process to pursue the proper pathways to serving individuals in the community, identifying those who need to be treated medically, and making transitions among locations and personnel.

Enter ExDS. The term excited delirium was first used in medical literature in 1985 to describe a subset of patients who used cocaine and exhibited symptoms of psychosis before suddenly dying.3 Since the widespread illicit distribution of cocaine in North America during the 1980s and 1990s, the psychoactive drug landscape has become much more complex. Synthetic stimulants, some modeled after therapeutic compounds used in psychiatry and others modified from natural extracts, have toxicity profiles that mimic cocaine with additional effects. Synthetic cannabinoids cause perceptual disturbances and have stimulatory and autonomic effects that overlap with other classes of compounds. Adding to the complexity of patient care is the reality that substances may be used in combination, with ethanol frequently added to the mix.

Clinical presentations involving agitation and other abnormal manifestations of behavior led frontline medical personnel to recognize that patients with the most severe intoxication syndromes had some features in common. Because their behavior and the circumstances in which it was encountered precluded having any other medically relevant information apart from what could be observed at a distance, patients were categorized as having ExDS as a way to describe the problem. Adopting that nosology led quickly and logically to practice pathways involving the conceptualization of such individuals with the features listed in the Table as patients with potentially life-threatening intoxications. The possibility of other underlying contributory medical conditions exists, but the presence of intoxicating substances—whether older substances such as ethanol, cocaine, and phencyclidine, or new ones such as bath salts and K2—is the sine qua non of ExDS.

Insistence was made that patients go for emergency care and, since the effects of substances are potent and the professionals insisting rarely have the interviewing skills of Bell, the result may be physical conflict. In order to minimize that conflict and provide care as rapidly as possible, emergency medical professionals seek the safest, fastest-acting sedative agents available. Benzodiazepines are still used, but the onset of action of a high dose (2-5 mg/kg) of intramuscular ketamine is faster. As the circumstances of uncontrolled conflict can be more dangerous in the prehospital setting than in the more controlled environs of an emergency department or intensive care unit, ketamine has been used by emergency medical services teams with greater frequency. Initially, intubation rates with ketamine were higher compared with those involving midazolam or haloperidol in emergency settings.4 But practice has become more refined, and first responders of all kinds in many communities are expressing relief that there are fewer physical altercations and injuries in such patient encounters. Emergency medicine physicians and trauma surgeons see lower rates of damaging sequelae in the situations involving agitation and prehospital conflict.

Not a Diagnosis

So why do we have the conflict and debate about ExDS? ExDS is a label or a descriptor, not a diagnosis. In psychiatry, we label the syndromes we treat intentionally as disorders—not diagnoses—as a way to acknowledge profound lack of certainty about pathophysiology and the corresponding potential for a number of contributing causes and circumstances to give rise to the clusters of signs and symptoms. Previously, however, as reflected in DSM-I and DSM-II, we did tie our diagnostic impressions to ideas about underlying causes—predominantly psychoanalytic ones. Categorization was done without the means to objectively verify etiologies. The result was that, although many patients did improve with our treatments based on assumptions about the underlying psychopathophysiologic processes described by Sigmund Freud and his followers, many others were labeled as having conditions borne of mechanisms that did not apply to them. Progress has been made in understanding the many determinants of psychiatric illnesses, but recent iterations of the DSM have largely stuck to refining syndromic descriptions for the purpose of having a common clinical lexicon. Still, there are debates about where to draw diagnostic lines, and the problem of reification drives reductionistic errors that lead to substandard treatment of patients and blind ends in research.

ExDS has roots in a well-understood problem with a known pathophysiology and established protocols for lifesaving treatment. There are many sound reasons to believe that patients exposed to novel psychoactive substances with pharmacologic mechanisms that overlap with cocaine experience similar toxicity profiles and will benefit from similar treatments. Methamphetamines largely follow that logic. Because of their longer duration of action and tendency to cause more psychotic symptoms that persist,5 patients who have taken methamphetamines can be more difficult to evaluate and treat due to their resistance to standard engagement with clinicians. Due to the diagnostic uncertainty that such patients embody (along with those who use novel agents, including mixtures of intoxicants), a syndromic label for their presenting signs and symptoms was understandably constructed with concern for severe underlying pathophysiology in mind.1

What is not known, however, is whether patients with those presenting characteristics actually have dangerous metabolic derangements and impending cardiovascular collapse. Repeated rounds of circular logic in the ACEP white paper drove home the point that patients with ExDS were critically ill and needed to be treated with “aggressive chemical sedation” to prevent harm befalling them.1 A more recent review of ExDS is more balanced in acknowledging limitations of the construct and the clinical research underlying it,6 including the fact that only a small fraction of patients identified as having ExDS have documented acidosis on laboratory testing once blood can be drawn for analysis. The review pointed out that the ExDS criteria are highly subjective and inconsistently applied in practice. Although the original ACEP white paper authors recommended a minimum of 6 of the 10 criteria be present to warrant a diagnosis of ExDS, numerous case series about ExDS involved the presence of fewer than 6 and sometimes as few as 1. Still, the nosology remains in place.

Encountering Agitated Individuals

Further complicating matters is the reality that physicians are not the professionals directly observing the prehospital features of most patients who may meet criteria for ExDS. Patients get to emergency departments based upon the impressions of first responders. Even a brief reading of the list of ExDS criteria makes it clear that they are open to a wide range of interpretations by different observers. For this reason, we have placed the term agitation in italics throughout this article; even in hospitals, use of this word is far from precise and may therefore spark a variety of responses depending upon the situation and the characteristics of the individuals involved.

Efforts have been made to educate those who encounter agitated individuals in the community with attention to the importance of identifying ExDS. With subjective criteria, however, and all the complex factors that bias observation and human decision-making, many patients who do not have a dangerous medical condition may be treated aggressively as if they do. As we have learned in psychiatry with examples such as the overdiagnosis of persons of color with schizophrenia compared with the related condition of bipolar disorder, criteria-based entities such as ExDS are subject to the influences of classism, racism, and a host of systemic biases. In the first large study reviewing psychiatric outcomes after prehospital use of ketamine for severe agitation, the only patient characteristic that predicted treatment with that agent over benzodiazepines was being Black.7 For these reasons—and citing that the use of the term ExDS may justify “law enforcement use of excessive force”—the American Medical Association (AMA) recently “oppose[d] the use of excited delirium as a medical diagnosis” after approving a position statement on the matter in December 2020.8

At a Crossroads

Emergency medicine finds itself at the crossroads with ExDS that psychiatry faced after coming to appreciate the pitfalls of DSM-II. Working with the limitations of a list of diagnostic criteria tied to assumptions about underlying pathophysiology, the specialty has employed sedative medications, including ketamine, at increasing rates. Reification has narrowed clinical thinking, and prehospital and emergency department protocols have sometimes become more clinician-centered than patient-centered. Many symptoms of ExDS emerge not only from within the psychosomatic experience of patients, but as a result of how patients are treated by those who first encounter them. It is important to note that a large review of patients identified with ExDS indicated that, despite concerns about an underlying deadly pathophysiology, the patients who die are the patients who are restrained.9

Some emergency medicine practitioners aver that fast-acting sedatives such as ketamine are vital tools that should be used more liberally to minimize injuries and decrease duration of physical restraint. Restraint should be avoided as it may exacerbate the metabolic acidosis feared to be a driver of lethality.10 Errors have been made as these issues have converged and the drug landscape has evolved—many of them referable to problems in the relationships between and among medicine, police, and the communities served. Noting that the world watched on video a man who clearly did not have ExDS die as a result of protracted restraint by police places the problem in stark relief. Even an expert clinical observer would not ascribe the necessary number of criteria for the syndrome, especially because George Floyd demonstrated a period of lucid calm that clearly did not change until after police engagement. Yet, an officer on the job for less than 2 weeks said of the prone Mr Floyd, “I just worry about the excited delirium or whatever.”11 That utterance emphasizes the imprecision of the ExDS criteria and the potential consequences of its misapplication in hot moments fraught with unconscious bias in human interaction.

Differing opinions representing substantive perspectives about ExDS persist, with corresponding recommendations for research and treatment approaches (not unlike the ongoing debates we have in psychiatry about autism spectrum conditions, bipolar phenomena in youth, and somatic symptom disorders). What should be appreciated and embraced is where this hotly debated entity fits within the range of human experiences. ExDS points to the potential existence of a life-threatening condition that is quite rare. Many of its criteria, however, line up with a wide variety of behavioral presentations that do not correlate with critical illness. Depending upon which criteria are observed in a given individual and how those criteria are ascertained, imminent danger to the individual in question and/or those in the vicinity may not be nearly as high as the ExDS literature would suggest.

On any given day in North America, a lot of individuals exhibit concerning behavior (Figure). A small fraction of those individuals use drugs that, under particular circumstances of interpersonal conflict, ambient temperature, and underlying physical vulnerability, may produce severe somatic illness.

Figuring out the best way to approach such individuals in the uncontrolled environments where they are encountered involves practical knowledge of the landscape of all the potential causes of behavioral difficulty, not just the tiny subset of those who may have a serious condition with a verifiable corresponding pathophysiology. Ideally, the large numbers of individuals in crisis with other underlying causes would be treated astutely and humanely by first responders with an ability to intervene without further escalating the situation.

Skills can be taught that allow better care for individuals with histories of trauma, autism spectrum conditions, and psychiatric disorders. Prehospital professionals need more attention to their education in these areas compared with the narrow focus on potential physical acuity, including ExDS. Abandoning the term may acknowledge that its use has yielded considerably fewer benefits to medical progress compared with much greater harms both for individuals and for the relationship between medicine and society (not unlike psychiatry’s abandonment of the category of hysteria). As it stands, ExDS is not a valid clinical diagnosis; it is a loosely defined syndrome not directly tied to 1 verifiable pathophysiology. It is a label that points to a problem at the interface of medicine, law, and society that has a differential diagnosis of etiologies embedded within it that do not all have the same treatments. Those validated conditions need to be identified and managed properly for each patient who appears to have ExDS as the starting point for care.

Help More Than Harm

The path forward will need to address the many forms of systemic bias embedded in society and the structures of service professions in order to refine the work in a just way. Because of the dangers and uncertainties inherent to the prehospital setting, we must acknowledge that some larger fraction of individuals than those with genuine deadly toxicity underlying presentation with ExDS criteria will still need to be treated as if they might have it in order to save those who do. Finding our collective range of tolerance for an acceptable range of unavoidable error in the face of such complexity is a task that will involve, among other things, greater tolerance.

It is paramount that the procedures we adopt do not allow the perspectives of medicine to harm more individuals than are helped. Physicians tend to believe that the details of medical science are too difficult for the public we serve to comprehend and thus exclude laypeople from dialogue about the values and actions of our profession. We must resist using that idea as an excuse around the issue of ExDS and instead productively engage the society we serve about the experiences, needs, and expectations of patients and their advocates to help foster real understanding and repair relationships going forward.

Concluding Thoughts

In the end, the situation remains complex, and even widespread implementation of the aforementioned suggestions will not resolve all the potential problems. Outlier cases will still confound ideal management. Expectations for first responders, prehospital emergency medical personnel, and emergency physicians remain extremely high, and the threat of litigation looms behind each encounter. Emergency medicine regularly “hopes for the best but defends against the worst” in standard practice so that dangerous conditions are not missed and lives are saved. One study of patients who were intoxicated (most cases involved ethanol) demonstrated that 1% had a comorbid underlying critical illness that, had it not been identified, would almost certainly have resulted in a morbid or mortal outcome.12

As previously noted, patients who have such complex conditions do not always find their way to emergency departments for care calmly or without forceful assistance. The threat of violence in a society full of racial and class conflict adds enormous burden to the demands for rapid decision-making in situations that may or may not need to be handled aggressively to prevent harm and/or deliver sick patients to the services they need. Choosing to be less insistent with medical interventions in order to honor human dignity and autonomy may very well lead to more cases of missed diagnosis and poor physical outcome in exchange for attempts at more immediate improvements in psychological and social outcomes. The interface of these choices with protocols of law enforcement demonstrates clearly that asking police officers to function as clinicians is not the answer.

Psychiatry has an opportunity to be helpful. Recognizing the high intensity and acuity of the situations in question and not reacting merely with finger-pointing judgment of colleagues is an essential start. Using the parallels between our specialty’s difficulties with diagnostic uncertainty and evolution as well as lessons from historical nosological missteps to assist our emergency medicine colleagues may come next. We also have opportunities to point out how the behavior of individuals and groups often reflects systemic biases that appear to serve necessary purposes but unconsciously do harm. And we can make ourselves available to assist with reforms by educating other disciplines about the many conditions that manifest abnormal behavior with accompanying methods to distinguish among them and then respond in more nuanced, empathic, and therapeutic fashion. Such collaboration opens up opportunities to remind specialists in emergency and critical care that, for all the advantages of efficiency offered by rapidly administered parenteral ketamine, the psychiatric consequences of having a drug given without consent that causes profound dissociation may be severe—especially for those with a preexisting history of trauma.

Embedding psychiatrists, psychologists, social workers, and other colleagues with mental health expertise in the teams (police units and ambulance services) and locations (the streets and emergency departments) where patients first present with behavioral unrest may decrease this use of ketamine in favor of dialogue and partnership that both protects the public and gets patients the care they need. Early successes with a model that pairs mental health professionals with and/or replaces police depending upon the nature of the emergency situation have been realized in Oregon and Aurora, Colorado—the city in which Elijah McClain was wrongfully stopped, physically restrained, and injected with ketamine and subsequently died.

Like the AMA, the American Psychiatric Association has put forth a position statement about the term excited delirium.13 Before deciding on one’s opinions about what to make of ExDS and all the relevant historical, medical, ethical, and social concerns, we wish to make 1 direct recommendation for our readers: Call the medical director of emergency medical services in an urban catchment area and ask to ride an evening shift in an ambulance. There is much to see, much to learn, and much to be done.

Dr Rasimas is medical director of consultation-liaison psychiatry at Hennepin County Medical Center in Minneapolis, Minnesota; professor of psychiatry at the University of Minnesota in Minneapolis; and associate professor of psychiatry and emergency medicine at the Penn State Milton S. Hershey Medical Center in Hershey, Pennsylvania. Ms Rasimas graduated from the University of Virginia with a degree in neuroscience and completed a research fellowship at the National Institutes of Health in Bethesda, Maryland. She received her master’s degree in public health from the University of Minnesota in which she focused on the problems of gun violence and suicide. She is currently a science writer in Saint Paul, Minnesota.


1. ACEP Excited Delirium Task Force. White Paper Report on Excited Delirium Syndrome. American College of Emergency Physicians; 2009.

2. Kraines SH. Bell’s mania. Am J Psychiatry. 1934;91(1):29-40.

3. Wetli CV, Fishbain DA. Cocaine-induced psychosis and sudden death in recreational cocaine users. J Forensic Sci. 1985;30(3):873-880.

4. Olives TD, Nystrom PC, Cole JB, Dodd KW, Ho JD. Intubation of profoundly agitated patients treated with prehospital ketamine. Prehosp Disaster Med. 2016;31(6):593-602.

5. Rusyniak DE. Neurologic manifestations of chronic methamphetamine abuse. Psychiatr Clin North Am. 2013;36(2):261-275.

6. Gonin P, Beysard N, Yersin B, Carron PN. Excited delirium: a systematic review. Acad Emerg Med. 2018;25(5):552-565.

7. Lebin JA, Akhavan AR, Hippe DS, et al. Psychiatric outcomes of patients with severe agitation following administration of prehospital ketamine. Acad Emerg Med. 2019;26(8):889-896.

8. Pharmacological intervention for agitated individuals in the out-of-hospital setting H-130.932. American Medical Association. Updated 2021. Accessed January 4, 2022.

9. Strömmer EMF, Leith W, Zeegers MP, Freeman MD. The role of restraint in fatal excited delirium: a research synthesis and pooled analysis. Forensic Sci Med Pathol. 2020;16(4):680-692.

10. Hick JL, Smith SW, Lynch MT. Metabolic acidosis in restraint-associated cardiac arrest: a case series. Acad Emerg Med. 1999;6(3):239-243.

11. Neuman S. Transcripts of police body cams show Floyd pleaded 20 times that he couldn’t breathe. WAMU 88.5 American University Radio. July 8, 2020. Accessed January 4, 2022.

12. Klein LR, Cole JB, Driver BE, Battista C, Jelinek R, Martel ML. Unsuspected critical illness among emergency department patients presenting for acute alcohol intoxication. Ann Emerg Med. 2018;71(3):279-288.

13. Council on Psychiatry and Law. Position statement on concerns about use of the term “excited delirium” and appropriate medical management in out-of-hospital contexts. American Psychiatric Association; 2020. ❒