ORIGINAL CONTRIBUTION violence, emergency department
Aggression Directed Toward Emergency Department Staff at a University Teaching Hospital Study objective: To determine the scope and magnitude of patient and visitor aggression directed toward emergency department staff. Design: One-year retrospective review of university police log records and ED staff incident reports. Setting: Medium-sized, urban, noncounty, university Level I teaching hospital treating approximately 40,000 ED patients annually Type of participants: All violent incidents involving patients~visitors and ED staff that triggered a police response to the ED area were included in the study. Interventions: None. Measurements and main results: All ED violent episodes were recorded and categorized by shift, type of incident, type of police response, perpetrator, and site of incident. It was found that police responded to the ED nearly twice daily; the night shift had 32% of the cases with only 13% of the patient volume; custody and medical psychiatric clearance patients accounted for 40% of the cases; more than 20% of incidents occurred in the waiting room; and 4.2% of the incidents represented a significant threat to ED staff. Conclusion: ED violence is a significant and under-reported problem at our medium-sized university teaching hospital. These data are useful in objectively quantifying the scope of violence in our institution, and they underscore the potential risks to emergency patients, visitors, and staff. There is an acute need for additional studies in other settings so that appropriate and cost-effective security recommendations can be formulated. [Pane GA, Winiarski AM, Salness KA: Aggression directed toward emergency department staff at a university teaching hospital. Ann Emerg Med March 1991;20:283-286.]
Gregg A Pane, MD, MPA, FACEP* Orange, California Anne M Winiarski, MDt Denver, Colorado Kym A Salness, MD, FACEP* Orange, California From the Division of Emergency Medicine, University of California, Irvine;* and the Denver Affiliated Emergency Medicine Residency Program, Denver, Colorado.t Received for publication June 26, 1990. Revision received October 15, 1990. Accepted for publication November 23, 1990. Presented at the Third International Conference on Emergency Medicine in Toronto, Ontario, Canada, June 1990. Address for reprints: Gregg A Pane, MD, MPA, FACER UC Irvine Medical Center, 101 The City Drive, Route 128, Orange, California 92668.
INTRODUCTION Despite widespread concern with the problem of violence in the emergency department and its increasing prominence in the media, there is little objective information addressing this topic in the literature. Because of this lack of information, the full extent and magnitude of violence in the ED are difficult to determine. Most studies on ED violence have been published in the British literature. 1-4 By comparison, there has been little study of the magnitude or incidence of violence directed against emergency personnel in the United States. It follows that specific security guidelines to address the problem are essentially nonexistent. Because of the high profile of this issue, the lack of objective data, and its substantial public policy implications, we conducted a retrospective review of police records at our institution to obtain specific data on the incidence and type of ED violence. This information would be useful in formulating security recommendations and in beginning to fill the data gap in the medical literature. MATERIALS A N D METHODS The University of California Irvine (UCI) Medical Center is 493-bed, noncounty, university teaching hospital and Level I trauma center that treats approximately 40,000 emergency patients annually. At the time of the study, the ED served as the primary screening facility for all county jail
94/283
Annals of Emergency Medicine
20:3 March 1991
AGGRESSION Pane, Winiarski & Salness
patients and for many psychiatric patients from .Orange County who required medical clearance. There were no security officers stationed in the ED on a regular basis. The medical center serves a primarily young population, with more than 60% of the patients being less than 30 years old. More than half of the patients are Hispanic, and approximately 75% are either uninsured or covered by a public aid program (Medi-Cal or County Indigent Medical Services). A p p r o x i m a t e l y two thirds of the patients have a family income of less than $10,000. Orange County has 37 hospitals and a population of 2.3 million. A one-year retrospective review of UCI police log records from July 1986 through July 1987 was conducted to determine the number and type of security responses to the ED for incidents of aggression or violence. The police log denotes each episode in which the UCI police were summoned to respond to the ED for some type of disturbance or violent act. All events are recorded chronologically, with a brief description of the type of event and the location. The log does not include ED violent or aggressive incidents in which the staff did not choose to summon the police. All staff and outside agency requests for UCI police assistance in the ED for visitor or patient aggression directed toward staff were recorded and categorized by shift, type of incident, type of police response, perpetrator, and site of incident. The UCI police code system assigns a number for each type of situation and disturbance. For example, numbers are assigned for disturbing the peace, suspicious behavior, being drunk in public, uncooperative patient, custody patient, and medical psychiatric clearance. From reviewing the many different codes recorded on the police log, three basic categories of incident were recognized as being the most significant, into which nearly all of the codes could be conveniently and inclusively grouped for greater clarity. The three major categories were general disturbance, custody patient, and medical psychiatric clearance. General disturbance included the codes for disturbing the peace, suspicious behavior, drunk in public, and uncooperative patient as well as a number of other closely related de20:3 March 1991
scriptions. The custody patient classification related to patients brought from the Orange County Jail for primary screening before incarceration or i n m a t e s w h o developed a c u t e medical problems. Generally, these patients were accompanied by a security officer from the jail. However, if there was a significant disturbance involving a custody patient, the UCI police would be summoned to assist. Medical clearance for psychiatric patients included patients brought by area police departments for medical evaluation after they were determined to be suicidal, homicidal, or gravely disabled and unable to care for themselves. These patients were screened for organic disease and were then referred to p s y c h i a t r y if no acute .problem was detected. This group also included patients from our outpatient psychiatric unit who were sent to the ED for medical evaluation as a protocol before being admitted to psychiatry. Police response to the incident was recorded as verbal advisement, crowd control, show of force, physical restraint, removal from premises, or active intervention. Active intervention (some c o m b i n a t i o n of arrest, search, restraint, and removal from premises) implied that there was a significant threat to ED staff (eg, weapon, physical abuse, severe verbal threat) as noted by completion of a separate full police report form in addition to the log record. Incident sites were divided into four regions: main ED t r e a t m e n t area, custody/psychiatry room, triage/waiting room, and ambulance ramp. All data were abstracted from police log records and report forms, categorized, and then compiled into tables for analysis. In addition, all official incident report forms dealing with violent behavior that were completed by ED staff and sent to the Risk Management Department were reviewed.
RESULTS The UCI police were summoned to respond to 686 episodes of ED violence during the one-year study period. Nearly 75% of the incidents occurred during the evening or night shift. In the overwhelming majority of cases, police were called by ED staff (Table 1). Sixty percent of the episodes involved general disturbances, Annals of Emergency Medicine
TABLE 1. Proportion of violent incidents per shift Shift Day Evening Night Total
No. of Incidents 175 293 218 686*
%
% of Total ED Volume
25.5 42.7 31,8
42.8 43.9 13.3
*In 560 incidents (816%), the call for a police response was initiated by ED staff.
TABLE 2. Incident sites Site Main treatmentarea Custody/psychiatryroom Triage/waitingroom Ambulance ramp
No. of Incidents
%
336 201 122 27
49.0 29,2 17.8 4.0
13% involved custody patients, and 26% involved medical clearance for psychiatry patients. In 67% of cases, the police response to aggressive or violent behavior consisted of verbal advisement, basic crowd control, or show of force. More than 25% of the incidents required physical restraint or removal from the premises (off of the grounds or to psychiatry). Active police i n t e r v e n t i o n was required in 4.2% of the cases that represented significant threats to staff and involved a weapon, physical abuse, or a serious verbal threat. No serious injuries to ED staff occurred. Patients were the most frequent perpetrators of violence (77.3%), and v i s i t o r s m a d e up the r e m a i n i n g 22.7%. Nearly 80% of the incidents occurred in the ED treatment area or custody/psychiatry room, with the remainder taking place in the triage/ waiting area or ambulance ramp (Table 2). There were only seven incident report forms relating to violent behavior completed by ED staff during the study period. For each of these seven incident reports, no corresponding notation in the police record was found. Most involved an unpleasant verbal exchange between staff and patient or family, for which the police were apparently not summoned.
DISCUSSION In a medium-sized (40,000 annual 284/95
AGGRESSION Pane, Winiarski & Salness
patient volume), urban, university t e a c h i n g h o s p i t a l ED, police responded to violent behavior an average of nearly twice daily during the one-year study period. Forty percent of the cases involved custody patients and medical clearance for psychiatric patients. Nearly one third of the incidents occurred during the night shift. There is substantial concern regarding the problem of aggression and violence directed toward staff in hospital EDs. This concern was reflected by a resolution calling for increased hospital security measures at the 1987 American Medical Association House of Delegates.S, 6 Despite this, there are very few studies addressing this topic in the medical literature. Much of the available evidence is anecdotal because most hospitals do not keep separate records for ED violence. Violence is thought to be increasing in EDs in Los Angeles County, although most of the data are nonscientific and unsubstantiated. Police respond to ED violence an average of five times daily at Martin Luther King Medical Center and approximately once a week at both Santa Monica Hospital and the Veterans Administration Hospital in West Los Angeles. 7 We found that 29 of the 686 cases (4.2%) r e p r e s e n t e d a s i g n i f i c a n t threat to ED staff and involved a weapon (nine cases), physical abuse (17 cases), or severe verbal threat (three cases). A national survey of US teaching hospital EDs with patient volumes of more than 40,000 annually found that 18% had at least one t h r e a t a m o n t h to staff w i t h a weapon, 43% had at least one physical attack a month on a staff member, and 46% confiscated at least one weapon a month from patients, s A British National Health Service survey of 3,000 employees found that 23% of ED p e r s o n n e l had been threatened with a weapon during the past year, by far the highest incidence of any hospital unit. 1 In Veterans Administration Medical Centers from 1982 through 1985, weapon c o n f i s c a t i o n i n c r e a s e d 51.7%, whereas there was only a 10% increase in the total number of visits and admissions.9 Nearly one third of all violent ED episodes at our facility were found to have occurred during the night shift, 96/285
between 11:00 PM and 7:00 AM. Fortythree percent occurred during the evening shift, between 3:00 PM and 11:00 PM. A six-month prospective study at the Edinburgh Royal Infirmary Accident and ED (annual volume, 65,000 patients) found that 37% of all violent incidents occurred between 10:00 PM and 2:00 AM.2 In a three-month study of ED violence at Alaska Native Medical Center, 60% of all violent patients were seen between midnight and noon. lo An evaluation of 200 consecutive patients at an urban general hospital ED in Ohio revealed that 65% of the night-time population had a current or past psychiatric illness, t t Although the UCI police log recorded 686 responses to the ED during the study, a corresponding review of official incident reports filled out by ED staff and sent to hospital risk management concerning violent behavior revealed that only seven had been completed. There are no other formal recording mechanisms for violent episodes in our ED, as is the case in most centers. The British National Health Service survey found that all types of violent incidents were grossly under-reported, as official recording occurred in only 35% of aggressive acts associated with minor injury, 31% associated with a weapons threat, and 18% associated with a verbal threat. 1 In another study, violent incidents in hospital EDs and psychiatric facilities were under-reported by at least fivefold. 9 There are several limitations to our study that should be noted. The first l i m i t a t i o n is inherent to its study design as a retrospective record review w i t h c o r r e s p o n d i n g difficulties of i n c o m p l e t e or illegible data. It was not possible to determine the exact nature of the violence or the full extent of the risk to staff. In addition, the findings are potentially generalizable only to similar-sized teaching i n s t i t u t i o n s that serve a comparable patient population. Nonetheless, police log records are fairly objective records of violent incidents. A prospective study on the topic of violence in the ED might present difficulties with under-reporting or over-reporting of episodes by staff. At the time of our study, no security officers were permanently stationed in the ED. A national survey noted that 62% of teaching hospital Annals of Emergency Medicine
EDs with patient volumes of more than 40,000 annually have 24-hour security presence. 8 ED staff initiated our police response in 81.6% of the cases, whereas an outside agency (usually a local police department) called ahead for assistance in the remaining 18.4%. ED staff summoned the police by either a telephone call or a panic button located in the triage area or nursing station that sent a direct signal to the hospital operator. No closed-circuit television monitors or metal detectors are used in our ED. More than 20% of the violent incidents recorded in our survey took place outside of the ED treatment area in the waiting room, at the triage desk, or on the ambulance ramp. It should be emphasized that visitors were the perpetrators of violence in 22.7% of all cases. An unpleasant waiting room environment, including a lack of distractions (eg, toys, television, magazines, telephones, music), a lack of access to refreshments, separation of family from patients, a lack of updating information, general noise and commotion, and a lack of patient understanding of triage systems (the sickest being evaluated first), has been identified as a significant factor in precipitating violent behavior.3,4,7,8j 2 Other important factors that appear to precipitate ED violence include overcrowding with large volumes of high-acuity patients with res u l t i n g l o n g w a i t i n g t i m e s ; an i n c r e a s e in the n u m b e r of drug abusers, alcoholics, mentally ill, and gang members; potential provocation of patients by fatigued and overworked staff; and the 24-hour opendoor policy of EDs and their frequent use as a medical clearance area for psychiatric patients.3,4,L8 To address violent and aggressive behavior in the ED, a multitiered approach is evolving that encompasses both prevention and management. Preventive measures include control of environmental factors that may provoke individuals with violent tendencies, 24-hour security presence and night access control, closedcircuit television monitoring, metal d e t e c t o r w e a p o n screening, staff training for recognition and proper de-escalation of previolent aggressive behavior, and establishment of behavioral emergency committees.l,3,4,8,13 20:3 March 1991
AGGRESSION Pane, Winiarski & Salness
Henry Ford Hospital reported on the results of their comprehensive 24-hour ED security system, which includes increased television monitoring and metal detector weapon screening. D u r i n g t h e f i r s t six months of the study, there were only two patient complaints, no one refused to proceed through the metal detector, and there were no legal challenges. A survey of ED and security staff showed strong support for the program and increased feelings of safety in the workplace. 13 In addition, emergency-alert systems, personal alarms, panic buttons, and direct telephone lines allow staff to notify security personnel of violent acts quickly. However, there have been few studies addressing which combination of security measures is appropriate and cost effective for a given ED. Our results show that ED violence is a noteworthy and significant concern at our institution. From the study we gained objective data that were extremely useful in developing recommendations for additional security measures. These recommendations were increased ED security presence, television monitors in tri-
20:3 March 1991
age and in the major treatment area, night access control through the ED entrance, triage redesign for improved protection of the triage nurse, and an improved waiting room environment, including comfortable chairs, m a g a z i n e s , p i c t u r e s , and plants. All of these recommendations have been accepted or implemented in our ED.
her help in preparing the manuscript.
CONCLUSION A retrospective review of police records and incident reports confirmed that ED violence is a significant and under-reported problem at our medium-sized university teaching hospital. The findings underscore the potential risk to emergency patients, visitors, and staff, and they point out the urgent need for additional studies in other settings. Acts of aggression and violence directed toward emergency personnel constitute a timely public policy issue of p a r a m o u n t concern, not only to emergency physicians but also to the general public. A state-of-the-art, comprehensive approach to the problem is necessary to ensure the safety of the nation's EDs.
5. Pinkney D: ER violence growing. A m Med News 1987;30:2,70.
The authors thank Sally Ann Adams for
Annals of Emergency Medicine
REFERENCES 1. Cullen EJ: Violence to staff in the health services. Br Health Safety Comm 1987;1-12. 2. Morgan M, Steedman DJ: Violence and the accident and emergency department:Health Bull 1985;4316: 278-282. 3. Cardwell 8: Aggression management: Violence in accident and emergency departments, part 1(2). Nurs Times 1984;80:32-34. 4. Wright B: Hostility in accident and emergency departments. Nurs Mirror 1985;16h42-44.
6. Lax hospital security threatens ER staff. Calif Phys 1989;6:25 26. 7. Kowsky K: A d m i t t i n g a health hazard: Violence plagues emergency rooms. Daily Breeze March 29, 1987, A1. 8. Lavoie F, Carter G, Danzl D, et al: Emergency department violence in US teaching hospitals. Ann Emerg Med 1988;17:1227-1233. 9. Drummond D, Sparr L, Gordon G: Hospital violence reduction among high-risk patients. JAMA 1989;261: 2531-2534. 10. Nighswander T: The demography and health consequences of violence in an emergency room setting. Alaska Med t982;24:7-10. 11. Summers W, Rund P, Levin M: Psychiatric illness in a general urban ER: Daytime vs nighttime population. ] Clin Psychiatry 1979~40:340-343. 12. Levenson IL: Dealing with the violent patient. Post grad Med 1985;78:329-335. 13. Thompson B, Nunn J, Kramer T, et ah Disarming the department Weapon screening and improved security to create a safer ED environment. Ann Emerg Med 1988;17:419.
286/97
Aggression Directed Toward Emergency Department Staff at a University Teaching Hospital Study objective: To determine the scope and magnitude of patient and visitor aggression directed toward emergency department staff. Design: One-year retrospective review of university police log records and ED staff incident reports. Setting: Medium-sized, urban, noncounty, university Level I teaching hospital treating approximately 40,000 ED patients annually Type of participants: All violent incidents involving patients~visitors and ED staff that triggered a police response to the ED area were included in the study. Interventions: None. Measurements and main results: All ED violent episodes were recorded and categorized by shift, type of incident, type of police response, perpetrator, and site of incident. It was found that police responded to the ED nearly twice daily; the night shift had 32% of the cases with only 13% of the patient volume; custody and medical psychiatric clearance patients accounted for 40% of the cases; more than 20% of incidents occurred in the waiting room; and 4.2% of the incidents represented a significant threat to ED staff. Conclusion: ED violence is a significant and under-reported problem at our medium-sized university teaching hospital. These data are useful in objectively quantifying the scope of violence in our institution, and they underscore the potential risks to emergency patients, visitors, and staff. There is an acute need for additional studies in other settings so that appropriate and cost-effective security recommendations can be formulated. [Pane GA, Winiarski AM, Salness KA: Aggression directed toward emergency department staff at a university teaching hospital. Ann Emerg Med March 1991;20:283-286.]
Gregg A Pane, MD, MPA, FACEP* Orange, California Anne M Winiarski, MDt Denver, Colorado Kym A Salness, MD, FACEP* Orange, California From the Division of Emergency Medicine, University of California, Irvine;* and the Denver Affiliated Emergency Medicine Residency Program, Denver, Colorado.t Received for publication June 26, 1990. Revision received October 15, 1990. Accepted for publication November 23, 1990. Presented at the Third International Conference on Emergency Medicine in Toronto, Ontario, Canada, June 1990. Address for reprints: Gregg A Pane, MD, MPA, FACER UC Irvine Medical Center, 101 The City Drive, Route 128, Orange, California 92668.
INTRODUCTION Despite widespread concern with the problem of violence in the emergency department and its increasing prominence in the media, there is little objective information addressing this topic in the literature. Because of this lack of information, the full extent and magnitude of violence in the ED are difficult to determine. Most studies on ED violence have been published in the British literature. 1-4 By comparison, there has been little study of the magnitude or incidence of violence directed against emergency personnel in the United States. It follows that specific security guidelines to address the problem are essentially nonexistent. Because of the high profile of this issue, the lack of objective data, and its substantial public policy implications, we conducted a retrospective review of police records at our institution to obtain specific data on the incidence and type of ED violence. This information would be useful in formulating security recommendations and in beginning to fill the data gap in the medical literature. MATERIALS A N D METHODS The University of California Irvine (UCI) Medical Center is 493-bed, noncounty, university teaching hospital and Level I trauma center that treats approximately 40,000 emergency patients annually. At the time of the study, the ED served as the primary screening facility for all county jail
94/283
Annals of Emergency Medicine
20:3 March 1991
AGGRESSION Pane, Winiarski & Salness
patients and for many psychiatric patients from .Orange County who required medical clearance. There were no security officers stationed in the ED on a regular basis. The medical center serves a primarily young population, with more than 60% of the patients being less than 30 years old. More than half of the patients are Hispanic, and approximately 75% are either uninsured or covered by a public aid program (Medi-Cal or County Indigent Medical Services). A p p r o x i m a t e l y two thirds of the patients have a family income of less than $10,000. Orange County has 37 hospitals and a population of 2.3 million. A one-year retrospective review of UCI police log records from July 1986 through July 1987 was conducted to determine the number and type of security responses to the ED for incidents of aggression or violence. The police log denotes each episode in which the UCI police were summoned to respond to the ED for some type of disturbance or violent act. All events are recorded chronologically, with a brief description of the type of event and the location. The log does not include ED violent or aggressive incidents in which the staff did not choose to summon the police. All staff and outside agency requests for UCI police assistance in the ED for visitor or patient aggression directed toward staff were recorded and categorized by shift, type of incident, type of police response, perpetrator, and site of incident. The UCI police code system assigns a number for each type of situation and disturbance. For example, numbers are assigned for disturbing the peace, suspicious behavior, being drunk in public, uncooperative patient, custody patient, and medical psychiatric clearance. From reviewing the many different codes recorded on the police log, three basic categories of incident were recognized as being the most significant, into which nearly all of the codes could be conveniently and inclusively grouped for greater clarity. The three major categories were general disturbance, custody patient, and medical psychiatric clearance. General disturbance included the codes for disturbing the peace, suspicious behavior, drunk in public, and uncooperative patient as well as a number of other closely related de20:3 March 1991
scriptions. The custody patient classification related to patients brought from the Orange County Jail for primary screening before incarceration or i n m a t e s w h o developed a c u t e medical problems. Generally, these patients were accompanied by a security officer from the jail. However, if there was a significant disturbance involving a custody patient, the UCI police would be summoned to assist. Medical clearance for psychiatric patients included patients brought by area police departments for medical evaluation after they were determined to be suicidal, homicidal, or gravely disabled and unable to care for themselves. These patients were screened for organic disease and were then referred to p s y c h i a t r y if no acute .problem was detected. This group also included patients from our outpatient psychiatric unit who were sent to the ED for medical evaluation as a protocol before being admitted to psychiatry. Police response to the incident was recorded as verbal advisement, crowd control, show of force, physical restraint, removal from premises, or active intervention. Active intervention (some c o m b i n a t i o n of arrest, search, restraint, and removal from premises) implied that there was a significant threat to ED staff (eg, weapon, physical abuse, severe verbal threat) as noted by completion of a separate full police report form in addition to the log record. Incident sites were divided into four regions: main ED t r e a t m e n t area, custody/psychiatry room, triage/waiting room, and ambulance ramp. All data were abstracted from police log records and report forms, categorized, and then compiled into tables for analysis. In addition, all official incident report forms dealing with violent behavior that were completed by ED staff and sent to the Risk Management Department were reviewed.
RESULTS The UCI police were summoned to respond to 686 episodes of ED violence during the one-year study period. Nearly 75% of the incidents occurred during the evening or night shift. In the overwhelming majority of cases, police were called by ED staff (Table 1). Sixty percent of the episodes involved general disturbances, Annals of Emergency Medicine
TABLE 1. Proportion of violent incidents per shift Shift Day Evening Night Total
No. of Incidents 175 293 218 686*
%
% of Total ED Volume
25.5 42.7 31,8
42.8 43.9 13.3
*In 560 incidents (816%), the call for a police response was initiated by ED staff.
TABLE 2. Incident sites Site Main treatmentarea Custody/psychiatryroom Triage/waitingroom Ambulance ramp
No. of Incidents
%
336 201 122 27
49.0 29,2 17.8 4.0
13% involved custody patients, and 26% involved medical clearance for psychiatry patients. In 67% of cases, the police response to aggressive or violent behavior consisted of verbal advisement, basic crowd control, or show of force. More than 25% of the incidents required physical restraint or removal from the premises (off of the grounds or to psychiatry). Active police i n t e r v e n t i o n was required in 4.2% of the cases that represented significant threats to staff and involved a weapon, physical abuse, or a serious verbal threat. No serious injuries to ED staff occurred. Patients were the most frequent perpetrators of violence (77.3%), and v i s i t o r s m a d e up the r e m a i n i n g 22.7%. Nearly 80% of the incidents occurred in the ED treatment area or custody/psychiatry room, with the remainder taking place in the triage/ waiting area or ambulance ramp (Table 2). There were only seven incident report forms relating to violent behavior completed by ED staff during the study period. For each of these seven incident reports, no corresponding notation in the police record was found. Most involved an unpleasant verbal exchange between staff and patient or family, for which the police were apparently not summoned.
DISCUSSION In a medium-sized (40,000 annual 284/95
AGGRESSION Pane, Winiarski & Salness
patient volume), urban, university t e a c h i n g h o s p i t a l ED, police responded to violent behavior an average of nearly twice daily during the one-year study period. Forty percent of the cases involved custody patients and medical clearance for psychiatric patients. Nearly one third of the incidents occurred during the night shift. There is substantial concern regarding the problem of aggression and violence directed toward staff in hospital EDs. This concern was reflected by a resolution calling for increased hospital security measures at the 1987 American Medical Association House of Delegates.S, 6 Despite this, there are very few studies addressing this topic in the medical literature. Much of the available evidence is anecdotal because most hospitals do not keep separate records for ED violence. Violence is thought to be increasing in EDs in Los Angeles County, although most of the data are nonscientific and unsubstantiated. Police respond to ED violence an average of five times daily at Martin Luther King Medical Center and approximately once a week at both Santa Monica Hospital and the Veterans Administration Hospital in West Los Angeles. 7 We found that 29 of the 686 cases (4.2%) r e p r e s e n t e d a s i g n i f i c a n t threat to ED staff and involved a weapon (nine cases), physical abuse (17 cases), or severe verbal threat (three cases). A national survey of US teaching hospital EDs with patient volumes of more than 40,000 annually found that 18% had at least one t h r e a t a m o n t h to staff w i t h a weapon, 43% had at least one physical attack a month on a staff member, and 46% confiscated at least one weapon a month from patients, s A British National Health Service survey of 3,000 employees found that 23% of ED p e r s o n n e l had been threatened with a weapon during the past year, by far the highest incidence of any hospital unit. 1 In Veterans Administration Medical Centers from 1982 through 1985, weapon c o n f i s c a t i o n i n c r e a s e d 51.7%, whereas there was only a 10% increase in the total number of visits and admissions.9 Nearly one third of all violent ED episodes at our facility were found to have occurred during the night shift, 96/285
between 11:00 PM and 7:00 AM. Fortythree percent occurred during the evening shift, between 3:00 PM and 11:00 PM. A six-month prospective study at the Edinburgh Royal Infirmary Accident and ED (annual volume, 65,000 patients) found that 37% of all violent incidents occurred between 10:00 PM and 2:00 AM.2 In a three-month study of ED violence at Alaska Native Medical Center, 60% of all violent patients were seen between midnight and noon. lo An evaluation of 200 consecutive patients at an urban general hospital ED in Ohio revealed that 65% of the night-time population had a current or past psychiatric illness, t t Although the UCI police log recorded 686 responses to the ED during the study, a corresponding review of official incident reports filled out by ED staff and sent to hospital risk management concerning violent behavior revealed that only seven had been completed. There are no other formal recording mechanisms for violent episodes in our ED, as is the case in most centers. The British National Health Service survey found that all types of violent incidents were grossly under-reported, as official recording occurred in only 35% of aggressive acts associated with minor injury, 31% associated with a weapons threat, and 18% associated with a verbal threat. 1 In another study, violent incidents in hospital EDs and psychiatric facilities were under-reported by at least fivefold. 9 There are several limitations to our study that should be noted. The first l i m i t a t i o n is inherent to its study design as a retrospective record review w i t h c o r r e s p o n d i n g difficulties of i n c o m p l e t e or illegible data. It was not possible to determine the exact nature of the violence or the full extent of the risk to staff. In addition, the findings are potentially generalizable only to similar-sized teaching i n s t i t u t i o n s that serve a comparable patient population. Nonetheless, police log records are fairly objective records of violent incidents. A prospective study on the topic of violence in the ED might present difficulties with under-reporting or over-reporting of episodes by staff. At the time of our study, no security officers were permanently stationed in the ED. A national survey noted that 62% of teaching hospital Annals of Emergency Medicine
EDs with patient volumes of more than 40,000 annually have 24-hour security presence. 8 ED staff initiated our police response in 81.6% of the cases, whereas an outside agency (usually a local police department) called ahead for assistance in the remaining 18.4%. ED staff summoned the police by either a telephone call or a panic button located in the triage area or nursing station that sent a direct signal to the hospital operator. No closed-circuit television monitors or metal detectors are used in our ED. More than 20% of the violent incidents recorded in our survey took place outside of the ED treatment area in the waiting room, at the triage desk, or on the ambulance ramp. It should be emphasized that visitors were the perpetrators of violence in 22.7% of all cases. An unpleasant waiting room environment, including a lack of distractions (eg, toys, television, magazines, telephones, music), a lack of access to refreshments, separation of family from patients, a lack of updating information, general noise and commotion, and a lack of patient understanding of triage systems (the sickest being evaluated first), has been identified as a significant factor in precipitating violent behavior.3,4,7,8j 2 Other important factors that appear to precipitate ED violence include overcrowding with large volumes of high-acuity patients with res u l t i n g l o n g w a i t i n g t i m e s ; an i n c r e a s e in the n u m b e r of drug abusers, alcoholics, mentally ill, and gang members; potential provocation of patients by fatigued and overworked staff; and the 24-hour opendoor policy of EDs and their frequent use as a medical clearance area for psychiatric patients.3,4,L8 To address violent and aggressive behavior in the ED, a multitiered approach is evolving that encompasses both prevention and management. Preventive measures include control of environmental factors that may provoke individuals with violent tendencies, 24-hour security presence and night access control, closedcircuit television monitoring, metal d e t e c t o r w e a p o n screening, staff training for recognition and proper de-escalation of previolent aggressive behavior, and establishment of behavioral emergency committees.l,3,4,8,13 20:3 March 1991
AGGRESSION Pane, Winiarski & Salness
Henry Ford Hospital reported on the results of their comprehensive 24-hour ED security system, which includes increased television monitoring and metal detector weapon screening. D u r i n g t h e f i r s t six months of the study, there were only two patient complaints, no one refused to proceed through the metal detector, and there were no legal challenges. A survey of ED and security staff showed strong support for the program and increased feelings of safety in the workplace. 13 In addition, emergency-alert systems, personal alarms, panic buttons, and direct telephone lines allow staff to notify security personnel of violent acts quickly. However, there have been few studies addressing which combination of security measures is appropriate and cost effective for a given ED. Our results show that ED violence is a noteworthy and significant concern at our institution. From the study we gained objective data that were extremely useful in developing recommendations for additional security measures. These recommendations were increased ED security presence, television monitors in tri-
20:3 March 1991
age and in the major treatment area, night access control through the ED entrance, triage redesign for improved protection of the triage nurse, and an improved waiting room environment, including comfortable chairs, m a g a z i n e s , p i c t u r e s , and plants. All of these recommendations have been accepted or implemented in our ED.
her help in preparing the manuscript.
CONCLUSION A retrospective review of police records and incident reports confirmed that ED violence is a significant and under-reported problem at our medium-sized university teaching hospital. The findings underscore the potential risk to emergency patients, visitors, and staff, and they point out the urgent need for additional studies in other settings. Acts of aggression and violence directed toward emergency personnel constitute a timely public policy issue of p a r a m o u n t concern, not only to emergency physicians but also to the general public. A state-of-the-art, comprehensive approach to the problem is necessary to ensure the safety of the nation's EDs.
5. Pinkney D: ER violence growing. A m Med News 1987;30:2,70.
The authors thank Sally Ann Adams for
Annals of Emergency Medicine
REFERENCES 1. Cullen EJ: Violence to staff in the health services. Br Health Safety Comm 1987;1-12. 2. Morgan M, Steedman DJ: Violence and the accident and emergency department:Health Bull 1985;4316: 278-282. 3. Cardwell 8: Aggression management: Violence in accident and emergency departments, part 1(2). Nurs Times 1984;80:32-34. 4. Wright B: Hostility in accident and emergency departments. Nurs Mirror 1985;16h42-44.
6. Lax hospital security threatens ER staff. Calif Phys 1989;6:25 26. 7. Kowsky K: A d m i t t i n g a health hazard: Violence plagues emergency rooms. Daily Breeze March 29, 1987, A1. 8. Lavoie F, Carter G, Danzl D, et al: Emergency department violence in US teaching hospitals. Ann Emerg Med 1988;17:1227-1233. 9. Drummond D, Sparr L, Gordon G: Hospital violence reduction among high-risk patients. JAMA 1989;261: 2531-2534. 10. Nighswander T: The demography and health consequences of violence in an emergency room setting. Alaska Med t982;24:7-10. 11. Summers W, Rund P, Levin M: Psychiatric illness in a general urban ER: Daytime vs nighttime population. ] Clin Psychiatry 1979~40:340-343. 12. Levenson IL: Dealing with the violent patient. Post grad Med 1985;78:329-335. 13. Thompson B, Nunn J, Kramer T, et ah Disarming the department Weapon screening and improved security to create a safer ED environment. Ann Emerg Med 1988;17:419.
286/97
