The psychiatric emergency room has become a major point of entry to acute psychiatric services for persons with severe mental illness, particularly those who are violent (1,2). Despite the frequency with which violent patients are seen in the psychiatric emergency service, controversy exists about the recommended management of acutely agitated and violent patients in this setting. Several recent review articles have suggested the following general approach: attempt to calm the patient by verbal means; make a differential diagnosis, that is, obtain a history, take vital signs, and do a physical examination and laboratory work, including a urine toxicology screen; treat the patient with an oral benzodiazepine, such as lorazepam, unless the patient is clearly psychotic; and use an oral or intramuscular antipsychotic only if the patient is determined to be psychotic (3,4).
As for medication, some authors recommend using intramuscular haloperidol alone on an emergency basis (5,6) and resisting the use of benzodiazepines alone or as an adjunct due to the risk of further decreasing inhibition, especially in treatment of patients with a personality disorder (5) or elderly patients (3). Others have recommended using a combination of benzodiazepines and antipsychotics if the cause of the violence is not clear (7).
Although each of these recommendations may have merit in certain situations, it is unknown how these approaches are applied in contemporary practice in psychiatric emergency rooms in the case of patients who are agitated and who appear to be imminently violent. The purpose of the study reported here was to survey current practices in a nationally representative sample of psychiatric emergency rooms to assess how acutely violent patients are managed. An additional aim of the survey was to learn whether clinicians believed it is possible to determine the cause of the violent behavior before a patient is medicated and whether they believed current treatments are effective in reducing violence.
Although no comprehensive national directory of psychiatric emergency rooms exists to our knowledge, one source of information on a large number of such facilities is the American Association for Emergency Psychiatry, a national organization with approximately 170 members. Participants in our study were the medical directors of 20 psychiatric emergency rooms located throughout the United States. These individuals were selected from the membership list of the American Association for Emergency Psychiatry using a stratified random sampling procedure.
First, the membership was divided into four geographic areas of the country—Northeast, North Central, South, and West. Using a table of random numbers, five psychiatric emergency rooms from each region were selected and the medical director was contacted (N=20). The medical directors were asked whether their emergency room was a receiving unit for acutely violent patients and whether they themselves directly participated in the evaluation and treatment of such patients. If either of these criteria were absent, a second emergency room medical director from the region was contacted. All of the individuals who met the inclusion criteria were invited to participate, and all of them agreed to do so after the purpose of the study was explained.
A structured interview was conducted by telephone with each medical director during March 1998. The interview covered a description of the psychiatric emergency service and a variety of aspects of the medical director's practices in the management of acutely violent patients—that is, patients who were agitated or who were deemed to be at immediate risk for becoming combative. (A copy of the interview manual is available from the authors.)
Characteristics of the emergency rooms
Fifteen emergency rooms were described by their medical directors as serving a predominantly inner-city population. Nine emergency rooms were in public or county hospitals, nine were in university hospitals, one was in a veterans hospital, and one was in a private hospital. Seventeen emergency rooms had licenses that allowed them to keep patients only up to 24 hours, at which point patients needed to be discharged or transferred. Three medical directors described their setting as located in a medical emergency room.
Seventeen medical directors indicated that it is very difficult to determine the etiology of violent behavior when previously unknown patients are brought into the emergency room in an agitated and violent condition. Although the emergency rooms had access to Breathalyzers and to laboratory testing that could measure blood alcohol levels and do serum or urine toxicology screens, many of the patients are too agitated to cooperate with such testing. Seventeen medical directors stated that the patients are usually so agitated that it is difficult to even obtain accurate vital signs to help with the differential diagnosis.
When the medical directors were asked what protocol they used most commonly for acutely violent patients, 14 said that these patients are usually put in restraints and medicated intramuscularly or intravenously and given a medical workup only after they are less agitated. Three said that such patients are usually medicated intramuscularly but are not placed in restraints, and three said that these patients are neither restrained nor given intramuscular medications. Almost all respondents (N=19) most frequently used four-point restraint when restraints were indicated.
Thirteen medical directors used the same acute medication regimen for all violent patients, regardless of the eventual diagnosis, whether it was functional psychosis, organic mental disorder, a substance-related disorder, or a personality disorder. The most common regimen, used by 11 medical directors, was haloperidol plus lorazepam, with or without benztropine. The second most common regimen, used by five medical directors, was droperidol, either alone (reported by four medical directors) or with lorazepam and diphenhydramine (reported by one medical director). Only one medical director used haloperidol plus benztropine without a benzodiazepine for all patients. Only three medical directors stated that violent patients are initially treated with only a benzodiazepine, such as lorazepam, when it is unknown whether the patient is psychotic.
As for preferred route of administration, 15 medical directors stated that the intramuscular route was most common. Two expressed a preference for using the intravenous route for medications when possible. One medical director stated that since he was located in a medical emergency room, some patients already have an IV line established by the time they are seen by psychiatric emergency staff. The other medical director stated that because he felt IV medications were so much more effective, he would call security personnel to restrain the patient and give the medications while the security personnel kept the patient's arm immobile. All 20 medical directors felt that their preferred medication regimen was effective for calming the violent patient, usually after one dose and always after one to two repeated doses.
Only three medical directors stated that the agitated patients treated in their emergency room will usually take medications orally and that restraints are rarely used. According to these medical directors, some of the factors allowing them to use these less coercive techniques included an integrated county system where most contacts are with people who are known to the system or who have case management protocols, a computerized system where information on patients is available within 30 to 60 seconds, a less violent patient population, and the availability of nurse clinicians who know the population.
To our knowledge, this study is the first to survey clinicians about their actual practices in the management of acutely violent patients in contemporary psychiatric emergency rooms. The sample size of 20 is small, which may limit the generalizability of the results. Nevertheless, the representativeness of the sample, its national scope, and the consistent findings suggest the value of further investigation with larger samples.
The results suggest that the strategies most frequently advocated in recent review articles for the assessment and management of violent patients are not generally applied by those responsible for the emergency management of acutely violent patients. When violent and agitated patients are brought to psychiatric emergency rooms, clinicians appear to place highest priority on prevention of patient and staff injuries by rapidly reducing violent behavior. Unless the patients are already well known to staff, clinicians most commonly manage acutely violent patients through restraints and intramuscular medications, typically a combination of neuroleptics and benzodiazepines, irrespective of diagnosis. Subsequently, potential causes of the violent behavior can be evaluated. Emergency room medical directors generally express the opinion that this medication regimen is effective for calming acutely violent patients and also permits referral to other levels of care within a 24-hour period.
One could argue that these practices involve risks of excessive coercion, overmedicating patients, and exacerbating underlying medical conditions. On the other hand, the clinical experience of practitioners suggests that these strategies rapidly ameliorate acute violence and thereby reduce the risk of injury.
The authors thank Judy N. Lam, Ph.D., for her assistance.
Dr. Binder is professor and Dr. McNiel is adjunct professor of psychology in the department of psychiatry at the University of California, San Francisco. Address correspondence to Dr. Binder at Langley Porter Psychiatric Institute, 401 Parnassus Avenue, San Francisco, California 94143-0984 (e-mail, firstname.lastname@example.org). Douglas H. Hughes, M.D., is editor of this column. Parts of this paper were presented at the annual meeting of the American Psychiatric Association held May 30–June 4, 1998, in Toronto.