- Email: [email protected]
Emergency medicine and police collaboration to prevent community violence
REVIEW
Emergency Medicine and Police Collaboration to Prevent Community Violence
From the Violence Research Group, University of Wales College of Medicine, Heath Park, Cardiff, United Kingdom. Received for publication February 3, 2000. Revisions received June 5, 2000, and September 28, 2000. Accepted for publication October 17, 2000. Address for reprints: Jonathan P. Shepherd, MSc, PhD, the Violence Research Group, University of Wales College of Medicine, Heath Park, Cardiff, CF14 4XY, United Kingdom; fax 00-44-2920-742442; E-mail [email protected]. Copyright © 2001 by the American College of Emergency Physicians. 0196-0644/2001/$35.00 + 0 47/1/114317 doi:10.1067/mem.2001.114317
Jonathan P. Shepherd, MSc, PhD
See related articles, p. 423 and p. 438. An important responsibility of emergency departments is the management of injuries sustained in assaults. Most assaults, including many causing serious injury, are not reported and not recorded by the police. This is important because police investigation and the conviction of offenders has a substantial deterrent effect and because information about the circumstances of intentional injury is key to prevention. Recent investigation of ED-police collaboration has shown that many of the injured, and ED staff want offenses to be reported but that there are attitudinal, logistic, and ethical-legal obstacles to achieving this. Organized joint efforts by emergency medicine personnel and police departments, on the basis of a sound legal and ethical framework to protect the rights of both victims and offenders, should deter more violent offenders and would-be violent offenders. They also provide the police with unique aggregate, nonconfidential information that is of substantial help in tackling violence. ED data can be used to measure and refine violence prevention initiatives and are being developed as the basis of a new, independent measure of police performance. Strategies, practical ideas to overcome obstacles, and directions for future research are suggested. [Shepherd JP. Emergency medicine and police collaboration to prevent community violence. Ann Emerg Med. October 2001;38:430-437.] INTRODUCTION
Emergency department personnel interface not only with medical services but also with local communities. They have extensive opportunities to work with local public services, of which police services are most important in maintaining law and order and preventing violence. In the context of violence as an important health issue, this article summarizes work that demonstrates the substan-
4 3 0
ANNALS OF EMERGENCY MEDICINE
38:4 OCTOBER 2001
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
tial extent to which violence, including violence that results in ED treatment, is not reported to the police; states why reporting is important; lists the obstacles to reporting; and identifies strategies and practical ideas about how obstacles might be overcome to the benefit of both communities and victims. It is concerned primarily with violence in which adults are injured. In setting out directions for future research, it identifies the need for evaluations of interventions that focus on both injury outcomes and process outcomes so that the benefits of ED-police collaboration can be maximized.
laws prevent violence and has been influential in improving criminal justice responses to handgun violence, alcohol-impaired driving, domestic violence, and child abuse. Although it might be assumed that criminal laws are built on good evidence of effectiveness, few have been quantitatively evaluated by using methods commonplace in medicine. The reasons for this include more reliance on qualitative approaches in legal and social sciences and reliance on incomplete and unreliable police crime statistics as measures of violent offenses.12 L O W P O L I C E R E P O R T I N G R AT E S F O R V I O L E N C E
I M P O R TA N C E O F V I O L E N C E A S A H E A LT H I S S U E
The importance of violence as a health issue is now well established. It is an important cause of death and physical and psychologic morbidity.1-3 Even in countries like the United Kingdom, which have strict weapon control laws, violence affects health substantially, particularly in terms of the facial deformity caused by blunt trauma.4,5 Leaving aside the injuries that violence produces, fear of crime can have a paralyzing effect on communities and can reduce quality of life for individual citizens.6 In turn, violence has been recognized as a public health problem, partly in the hope that, “It would be better to reduce violence through the positive aim of promoting health rather than through the negative aims of retribution, deterrence and incapacitation.”1,7 However, on their own, health, community safety, or criminal justice solutions to violence are not enough. There is now a consensus internationally that the key to community violence prevention is interagency coordination at the local level.8 Health services-criminal justice collaboration is important not only at the operational level but at the policy level as well. One of the most interesting recent developments in medicine is scrutiny of the effect of criminal laws by epidemiologists. It has been shown, for example, that accidental shooting deaths of children can be reduced by laws requiring gun owners to keep firearms locked away.9 An evaluation of state laws that restrict handgun purchase by persons with criminal records has demonstrated decreases in violent offenses committed by would-be purchasers who were denied handgun purchase.10 Targeted police work has been found to prevent alcohol-related violence in Australia by increasing deterrence.11 Studies like these suggest that law enforcement can produce changes in behavior. This health perspective of the criminal justice system is of great strategic importance. It provides a robust way of finding out whether criminal justice systems and criminal
OCTOBER 2001
38:4
ANNALS OF EMERGENCY MEDICINE
Another important reason to develop health servicescriminal justice partnerships is that EDs detect far more assaults in the community than are recorded by the police.12,13 Cross-referencing ED and police data concerning community violence has not been done systematically on a national basis, but studies in regional centers in the United Kingdom have shown that overall, only about 25% to 50% of offenses that led to ED treatment appeared in police records.12,13 This proportion is consistent with the findings of biennial British crime surveys, which compare householders’ accounts of crime with police records.14 In a UK study, assaults on men and assaults in bars and nightclubs and in the street that resulted in ED treatment were less likely to be recorded by the police than assaults on women and assaults in other locations.12 Differential rates of recording were also found according to time of day and day of week. This has obvious implications for the organization of community policing and law enforcement.12 In the United States, low rates of police recording-investigation of intimate partner violence have prompted legislation in some states to make reporting by health professionals compulsory.15 For firearm injury surveillance in the United States, although systems are being developed, systematic study of the extent to which there is correspondence between ED and police recording has yet to be done.16 A study of violence-related injuries treated in a sample of 31 US EDs in 1994 found that rates of injury inflicted by intimates (current or former spouse, boyfriend or girlfriend) were 4 times higher than rates from the National Crime Victimisation Survey.17 Because the purpose of this study was to estimate the overall burdens that violence imposes on EDs and not primarily to compare ED data with criminal justice data, the reasons for lack of police recording were not investigated. It was assumed, however, that such major differences could be ascribed to methodologic factors. Encouragingly for any future measurement of vio-
4 3 1
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
lence with ED records, however, there was close correspondence between overall rates of intentional injury when compared with rates from the 1994 National Hospital Ambulatory Medical Care Survey.18 Successive editions of the International Victimisation Survey have demonstrated that substantial numbers of violent offenses are not included in official crime statistics. They have also revealed that there are differences in rates of victimization between countries.19 For example, the 1996 survey showed that rates of victimization were highest in England and Wales, the United States, Sweden, and Finland and lowest in the Netherlands, Austria, and Northern Ireland. In the 11 industrialized countries that took part in the survey, the median reporting rate of violent offenses was 39% (range 18% to 51%).19 Although it might be thought that it is predominantly assaults causing minor injury that are not reported, 2 UK studies have found that police recording could not be predicted on the basis of injury severity scores and that many assaults causing severe injury are not reported.20,21 REASONS TO INCREASE REPORTING
If EDs and law enforcement worked together to enhance reporting of crimes, this could deter potential offenders, provide the police with information about violence that is not available from any other source, and help repair the wider damage done to victims and communities, potentially including the effects of restorative-reparative justice. There is now strong evidence that the police and the criminal justice process have a deterrent effect. International comparisons of police and crime survey measures of violence have demonstrated negative correlations between the chances of conviction and levels of violence.22-25 In 1981, the assault rate in England and Wales, according to victim surveys, was slightly higher than in the United States, but by 1995, it was more than double.22 The same trend is evident for other categories of violence. For murder, in 1985, rates in the United States were 8.7 times those of England and Wales, but by 1991, the excess risk had narrowed to 5.7 times. Over the same period, the proportion of assaults reported to the police in the United Kingdom did not change significantly, but improvements in recording rates for violence in the United States (63% recorded in 1981 to 93% recorded in 1995) were far greater than those noted in England and Wales (37% in 1981 to 46% in 1995). At the same time, there were obvious but opposite changes in rates of conviction for assault in the 2 countries.22 Between 1981 and 1995, conviction rates rose
4 3 2
steeply in the United States (from 0.16 to 0.44 per 1,000 population) but decreased markedly in England and Wales (from 1.12 to 0.61 per 1,000 population). A study of the effectiveness of urban closed-circuit television surveillance in Wales produced evidence consistent with this: only negative relationships were found between levels of community violence measured in EDs and levels according to police data.26 A recent authoritative review of research on criminal deterrence concluded that there had been negative correlations between the likelihood of conviction and crime rates studied over 10 years in England, the United States, and Sweden and over 15 years in England and the United States.25 This is consistent with previous research.27 Statistical associations between the severity of punishment and crime rates are much weaker than associations between certainty of punishment and crime rates and generally have not reached statistical significance.25 Although attempts have been made to control for influences on crime rates other than deterrence,28 the methods used have not been reliable enough to draw reliable conclusions.25 The principal source of quasiexperimental studies of deterrence is research on drunk driving.29,30 It has been concluded that, “There is considerable evidence that increasing the actual certainty of punishment for drunk driving in ways that also ensure adequate publicity of certainty of consequences can effect reductions in drunk driving.”31 Overall, the published evidence indicates that when would-be offenders perceive substantial risk of punishment, many are deterred from offending.25 This suggests that on the basis of a sound ethical framework, an important objective for primary and secondary injury prevention should be increasing the chances of punishment of violent offenders. This fits with the traditional public health strategy of targeting the agent of injury, in this case the offender or potential offender. Although treating violence as a public health issue is not new,1,7 formal collaboration between health and criminal justice services to maximize the chances of arresting and convicting violent offenders as a means of injury prevention has yet to be developed. An obvious application of this approach, given the low rates of recording by the police of offenses that result in ED treatment, is for EDs to facilitate and encourage increased rates of police reporting by or on behalf of the injured. The second main reason to improve reporting is that the circumstances of a great deal of community violence are not known to the police. Although it is often assumed
ANNALS OF EMERGENCY MEDICINE
38:4 OCTOBER 2001
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
that the police know about where injuries are sustained and which weapons have been used, this is often not the case.12,32 Just because police are often in EDs does not mean that they record and investigate all or even the majority of offenses. Information that can easily be obtained by ED personnel from victims33 is therefore often not available to the police and may be of substantial importance in targeting police resources to prevent violence or to arrest offenders. In cases with victims who do not want to report, this information can be provided by EDs in an aggregate, nonconfidential form. For example, because in the United Kingdom only about 1 in 9 assaults in licensed premises that result in ED treatment are recorded by the police,12 information from EDs can identify the bars and nightclubs that are the focus for violence. Similarly, EDs can provide unique information about the locations of work-, school-, and drug-related violence. In the United Kingdom, this ED perspective has even led to the identification and modification of a previously unrecognized weapon category, bar glassware, which is, in fact, the most frequently used weapon in the United Kingdom.34 Beer glasses in the United Kingdom are now manufactured almost exclusively of toughened glass. A recent randomized controlled trial shows that increasing the impact resistance of bar glasses reduces the risk of injury.35 If it had not been for ED-derived information, this advance would not have been made. It seems obvious that violence prevention should take account of risk of injury, as well as offenses reported to the police, but this does not seem to have been appreciated. Furthermore, a zero-tolerance policy demands that attention is paid to minor, as well as life-threatening, injuries. Increased reporting may not just be good for the community but also for patients because assaults are not usually isolated incidents in the lives of either victims or assailants. About half of assault victims treated in the ED have been assaulted before, and most violent offenses are committed by a small proportion of offenders.4,36,37 In a Philadelphia cohort study, it was found that 6% of male offenders (18% of all offenders) accounted for 52% of all arrests of juveniles. This was even more noticeable for violent offenses: 6% of chronic offenders accounted for 69% of all aggravated assaults, 71% of homicides, 73% of forcible rapes, and 82% of robberies.36,37 In London, similar results have been obtained: about 6% of male offenders (17% of all offenders) accounted for about 50% of all convictions.37 In Sweden, offending is even more concentrated.38 These data help to explain the high risk of repeat victimization. In a study in Kansas City of domestic homi-
OCTOBER 2001
38:4
ANNALS OF EMERGENCY MEDICINE
cide, a substantial proportion of victims were found to have attended local EDs for treatment of injuries sustained previously in cases of domestic violence.39 A collaborative approach also helps to safeguard the social rights of patients, particularly the right to privacy, and to prompt recognition and consideration of wider circumstances, expectations, and needs.40 Because having somewhere safe to live is essential to physical and psychologic well-being, rehousing on a permanent or temporary basis and increasing levels of security through, for example, lock or window replacement is important. Links between health and victim support services help to safeguard the dignity and social rehabilitation of the injured, as well as providing help and advice with claims for compensation. Several barriers to reporting have been identified: these may be categorized as attitudinal, logistic, ethical, and legal. A study of police-ED interaction carried out in 5 EDs in Wales, United Kingdom, showed that barriers to police reporting were, in decreasing order of importance, perceived to be the following: patient confidentiality, maintaining a neutral stance in relation to issues of blame, and protecting the patient.21 On the other hand, reasons for reporting, in decreasing order of importance, were perceived to be the following: reducing risk to others, severe injury, use of a weapon, and vulnerability of the injured. The skepticism with which the proposal that emergency physicians in the United Kingdom should be more proactive in reporting community violence to the police was met illustrates some of the attitudinal obstacles to the adoption of a more coordinated interagency approach to tackling violence.41 It was argued that violence prevention should not be medicalized because this would let official crime prevention agencies off the hook and divert some of the blame for violence to physicians. Working with the police to increase the chances of violent offenders being investigated was attacked as “unacceptable medical paternalism.”41 Trends toward partnership between public services to solve the problem of offending make these objections seem less appropriate now than previously.8 In contrast, in one study, half of emergency physicians, nurses, and clinical and managerial staff believed that they should have a role in victim protection and support, the detection of crime, and community crime prevention.21 Fewer obstacles have been identified in relation to attitudes of the police. In Wales, the police generally understood the ethical constraints on ED staff but wanted more violence to be reported. However, some police officers found it hard to come to terms with the reality of so much unreported violence and of the extra work this might pro-
4 3 3
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
vide.21 The need to prevent violence and protect patients and staff in EDs (eg, by ensuring that weapons are not brought into EDs) may divert police attention away from undetected community violence. The way the police respond to a report can influence whether an offense is recorded as well. The injured may be categorized as offenders, evidence may not be deemed to be sufficient, the victim may be unwilling to make a formal complaint, or the incident may be dismissed as a “family” affair.13 Victims’ attitudes are also important because the burden for supporting the prosecution process falls predominantly on them at every stage,13,42 including in the ED.21 The injured often do not go to the police because they cannot identify their assailant, they are marginalized as a result of alcohol or substance abuse, they are habituated to violence, they are unwilling to have their own conduct scrutinized, they are fearful of reprisals, or they have an ongoing relationship with their assailant.13 In this context, it is not surprising that even cases of serious gun violence can go unreported. However, many patients expect that once their medical needs are addressed, ED staff will report offenses on their behalf or at least rapidly summon police help to allow them to report.21 Victims and those who accompany them are not just concerned about physical injuries but also about the identification, arrest, and prosecution of the offender; prompt notification of the police; and prompt collection of evidence. Particularly at this early stage, patients want justice and look to ED staff to help them achieve this.21 In the ED, logistic obstacles to reporting include lack of facilities for patients to report, lack of a software facility to record basic information about the circumstances of violence, the often exclusively health agenda of EDs, and lack of links or interfaces between EDs and other community services, such as the police.21,33 Immediate availability of a police officer, police politeness, and preexisting professional relationships with police officers influence whether ED staff report offenses.21 Case studies illustrate these obstacles (Figure 1). In one study, the bureaucracy associated with reporting serious offences to the police led to informal police tip offs.21 Interestingly, in this study, individual clinicians had far more experience of contacting police officers than individual police officers had dealing with ED staff: relative inexperience of ED work may therefore be a logistic obstacle for the police. Neither group, however, knew about the other’s formal guidance or protocols. Conversations with ED receptionists were often perfunctory and limited to the recording of demographic
4 3 4
information, cause of injury, and previous ED visits. A potential logistic problem is the often protracted nature of ED treatment, which in the Wales study involved waits of up to 5 hours for the patients surveyed in waiting areas, in treatment cubicles, and on trolleys. These waits nearly always reflected how busy emergency physicians were, but they do provide clear opportunities and time for other staff, particularly receptionists and nurses, to institute discussion of reporting and victim support issues.21 Overall, victims of assault make up about 2% to 5% of all new ED attenders,4 a proportion that may make violence seem a relatively minor problem, particularly in countries in which firearm injuries are comparatively rare. Late at night, this proportion is very much higher, but at this time, senior staff who are in a position to develop violence-prevention strategies with the police may not be present.21 Although links between EDs and other medical services are well established, active partnerships between EDs and nonmedical public services are relatively undeveloped. The sequence of events exemplified in the case study from Wales (Figure 1) was noted time and time again during observation of ED staff-patient interaction: the injured and those who accompanied them repeatedly presented ED staff
Figure 1.
Case study. A dazed 20-year-old man with an obvious facial laceration reported to an ED receptionist after being assaulted in a bar. Although he was incapable of speech because of his injuries, his girlfriend kept making comments about wanting police involvement between trying to control her tears and physically supporting her boyfriend. She told the receptionist, “the police don’t know yet,” and when an ambulance driver passed the reception area, she asked, “Is that a policeman?” The receptionist ignored all these comments. Later, when the man was lying on an ED trolley, the girlfriend asked the male nurse who was swabbing his face, “Can you take photographs in here?” His reply was, “No ... you had better take them tomorrow when all the swelling comes out.” The girlfriend then immediately stated that after the attack in the pub, “I asked them to ring the police .... It was completely unprovoked.” Later, when the receptionist was asked by a research worker why she had not said anything about the police, she said that, “If she asked me, I would have told her ... we can’t do anything. They have to tell the police themselves.” When the research worker pointed out that people who had been attacked might not be aware of this, she agreed that, “If he hadn’t been assaulted before, he wouldn’t know the procedure.” Similarly, when the research worker asked the nurse why he had not responded more positively about taking photographs, he replied, “We haven’t got time to take photographs, and anyway, I wouldn’t want to go to court. It’s bad enough meeting patients in Tesco’s (a supermarket).” At the time, this male nurse appeared to have plenty of time: he was eating curry in the staff restroom. He stated that in relation to supporting victims, “we are taught not to be judgmental.”21
ANNALS OF EMERGENCY MEDICINE
38:4 OCTOBER 2001
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
with obvious cues to liaise with the police, but, usually under the guise of confidentiality, these were not acted on.21 Respecting the confidential nature of the physicianpatient relationship is of course of major importance, and apart from exceptional circumstances, it should be for patients to decide whether offences are reported. However, patients should be advised, in their medical interests and in the medical interests of others, to report offenses if there seems to be a risk of future harm. Importantly, the interests of patients and the wider community do not usually conflict.41 Conflict may arise, however, when the victim is a potential suspect. In the United Kingdom, physicians are advised to report offences, without the patients’ consent if necessary, in cases in which it appears that not doing so may hinder the detection, investigation, or prevention of a serious crime.21 Confidentiality can be an obstacle to responsible, ethical reporting if it is used by health professionals as a reason not to consider the needs of patients, communities, and the public services that serve them.21 Failure to consider these needs when treating a victim could, of course, lead to protection of offenders from prosecution. Of importance, however, is the possibility that unauthorized disclosure and a closer relationship between EDs and the criminal justice system might deter some people from seeking the medical help they need. Whether the knowledge that information may be disclosed actually stops people from seeking treatment is open to question, however.43 Few countries have laws designed to ensure the reporting of violent offenses by health professionals, although some US states have legislation to ensure that domestic violence is reported.15 The reasons for this include the sometimes fatal consequences of repeat violence and the threat that domestic violence poses to families and children. In the United Kingdom, reporting is left to the discretion of physicians. So far, however, on both sides of the Atlantic, there has been little or no requirement for ED staff to collect the information on which rational judgements about reporting can be made. In contrast to adult violence, in most countries, the investigation of child abuse has been the subject of legislation to ensure an appropriate multiagency approach. Surprisingly, although legislative steps that involve unauthorized disclosure of information have been taken to protect society from communicable disease, these have rarely been taken and enforced to protect society from violence, even though violence has become a leading cause of death, disfigurement, and disability.1-7 The lack of a legal framework to ensure that physicians make appropriate decisions about the reporting and
OCTOBER 2001
38:4
ANNALS OF EMERGENCY MEDICINE
investigation of adult violence may be an obstacle to protecting patients and communities. From an injury risk perspective, the focus in the United States on legislation concerning responses to domestic violence seems narrow. Very importantly, however, although legislation should protect the interests of citizens and communities, police responses to domestic violence can do more harm to victims than good.44-49 From an international standpoint, there are obviously potential abuses of closer collaboration between EDs and the police in countries with poor human rights records. For example, in France in 1944, attempts by occupying forces to oblige physicians to identify wounded members of the resistance prompted the President of the Ordre National des Medicines to remind all French physicians of their duty to preserve confidentiality. The text of his telegram still prefaces the “Guide d’Exercise Professionel.”50 In the United Kingdom, advocacy for a multiagency approach that breaks down barriers between services resulted in the Crime and Disorder Act of 1998.32,51,52 This act formally identified health authorities as partners with whom local government and the police must work to audit and tackle crime.52 I D E A S A B O U T H O W O B S TA C L E S C A N B E OVERCOME
Ideas about how these obstacles to reporting can be overcome are considered under the categories listed above. It is clear from the past research that some attitudes need to be changed. Education and training of ED personnel is needed to provide understanding of the wider needs of patients and the community in relation to deterrence, violence prevention, the causes of assault injury, the reporting of offenses, and the criminal justice process. Because the objective is ethical collaboration with the police, joint training seems important. An essential foundation of this training should be clear guidelines on circumstances in which reporting of offenses, without the patients’ consent if necessary, is justified. ED personnel (predominantly reception staff in the United Kingdom) who record details of offenses with suitably adapted software (Figure 2) also need training. Because this group has been shown to be in the best position to record information about the circumstances of violence and whether the assault has been reported to the police, training is necessary to enable them to do more than simply record administrative details.34 Minor changes to the organization of ED reception areas to facili-
4 3 5
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
tate the recording of what may be sensitive information may also be required. For example, patients and those who accompany them need to be able to have one-to-one conversations with receptionists away from lines or queues. Practical ways in which reporting in the ED can be facilitated and encouraged are listed in Figure 2. Because police crime statistics are known to be unreliable measures of community violence,12,19 assault injury rates are also a potentially invaluable measure of local, state, and national police performance and of the effectiveness of community violence prevention initiatives.53 The UK government has recently agreed that ED-derived data should be developed for this purpose.54 Further legislation in this area should be based on evidence of what works (see below) and should facilitate a coordinated approach to protect all adults and ensure access for victims to effective social and mental health interventions. It must also ensure that all serious violence is reported. DIRECTIONS FOR FUTURE RESEARCH
Although the strategies outlined in this article are rational in terms of assault injury prevention, it is vital that EDpolice partnerships on the basis of sound ethical principles are evaluated to find out whether they are effective. This is true both for existing and new partnerships. In
Figure 2.
Suggested intervention strategies.
particular, evaluations are needed of the use the police make of ED-derived reports of offenses and nonconfidential aggregate information about the circumstances of offenses and of joint interventions to target community violence. Evaluations should be focused not just on injury outcomes but also on process outcomes that local communities can adopt to ensure efficient and effective service coordination. In relation to the measurement of community violence and of police performance according to incidence and severity of assault injury, harmonization of police, and injury data sets is required. The effectiveness of laws mandating the reporting of domestic violence and, in the United Kingdom, of laws that include the health sector in community crime prevention also needs to be assessed. In summary, emergency physicians who treat victims of assaults should not assume that the police know about all or most violence. In all countries in which comparisons of police and crime survey data are possible, this has been shown not to be the case, even for serious offenses. Contacts with the ED are opportunities to improve reporting, deterrence, and violence prevention, as well as treatment and social care. While safeguarding the rights of both offenders and victims to treatment without fear of inappropriate disclosure, this is a rational step toward safer and more just communities in which the fear of crime is kept to a minimum. I thank Dr. Mike Maguire, School of Social and Administrative Studies, University of Cardiff, and Dr. David Farrington, Institute of Criminology, University of Cambridge, for their advice.
REFERENCES Strategies to increase rates of police reporting and investigation: • Free direct telephone lines from ED waiting areas to the police. • Incorporation of ED waiting areas into police patrols. • Asking all assault patients at the time of ED registration whether they would like assistance to report. • Reporting by ED staff on behalf of patients if they wish. • Mandatory reporting of serious violence. Strategies to target community policing more effectively: • Computerized collection of core data by admitting ED staff on the circumstances of community violence, particularly in the home, licensed premises, schools, and poor neighborhoods. • Regular disclosure to the police of nonconfidential aggregate information about assault time, weapons, assailants, and location. Strategies to measure community violence and police performance: • Development by the Centers for Disease Control and Prevention and National Institute of Justice of a core violence data set designed for ED use. • Inclusion of ED aggregate data in community audits of violence. • Publication of local, state, and national ED-derived information about community violence.
4 3 6
1.
Shepherd JP, Farrington DP. Assault as a public health problem. J R Soc Med. 1993;86:89-92.
2.
Rivara FP, Shepherd JP, Farrington DP. Gun laws. Lancet. 1993;342:809-810.
3. Joy D, Probert R, Bisson JI, et al. Post-traumatic stress reactions after injury. J Trauma. 2000;48:490-494. 4. Shepherd JP. Violent crime: an accident and emergency department perspective. Br J Criminology. 1990;30:289-305. 5. Bisson JI, Shepherd JP, Dhutia M. Psychological sequelae of facial trauma. J Trauma. 1997;43:496-500. 6.
Shepherd JP. Fear of crime. Lancet. 1990;335:796-797.
7. Moore MH. Public Health and criminal justice approaches to prevention. In: Tonry M, Farrington DP, eds. Building a Safer Society. Chicago, IL: University of Chicago Press; 1995:237-262. 8. Graham J, Ekblom P, Pease K, et al. Reducing Offending: An Assessment of Research Evidence on Ways of Dealing With Offending Behaviour. London, England: Home Office;1998. Research study 187. 9. Cummings P, Grossman DC, Rivara FP, et al. State gun safe storage laws and childhood mortality due to firearms. JAMA. 1997;278:1084-1086. 10. Wright M, Wintemute GJ, Rivara FP. Effectiveness of denial of handgun purchase to persons believed to be at high risk for firearm violence. Am J Public Health. 1999;89:88-90.
ANNALS OF EMERGENCY MEDICINE
38:4 OCTOBER 2001
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
11. Homel R, Hauritz M, Wortley A, et al. Preventing alcohol-related crime through community action: the surfers’ paradise. In: Homel R, ed. Policing for Prevention. Crime Prevention Studies 7. Monsey, New York, NY: Criminal Justice Press; 1997. 12. Shepherd JP, Shapland M, Scully C. Recording of violent offences by the police: an accident and emergency department perspective. Med Sci Law. 1989;29:251-257. 13. Clarkson C, Cretney A, Davis G, et al. Assaults: the relationship between seriousness, criminalisation and punishment. Criminal Law Rev. 1994;Jan:4-21. 14. Mirrlees-Black C, Budd T, Partridge S. British Crime Survey: England and Wales 1998. London, England: Home Office; 1998. 15. Cole TB, Flanigan A. What can we do about violence? JAMA. 1999;282:481-483. 16. Kellermann AL, Bartolomeos KK. Firearm injury surveillance at the local level: from data to action. Am J Prev Med. 1998;15:109-112. 17. Rand MR, Strom K. Violence-related Injuries Treated in Hospital Emergency Departments. Washington, DC: US Department of Justice; 1997. Publication NCJ-156921. 18. National Hospital Ambulatory Medical Care Survey (NHAMCS). Atlanta, GA: Centers for Disease Control; 1994. 19. Mayhew P, Van Dijk JJM. Criminal Victimisation in Eleven Industrialised Countries. Key Findings from the 1996 International Crime Victims Survey. Amsterdam, The Netherlands: WODC Ministry of Justice; 1997.
39. Wadman MC, Muelleman RL. Domestic violence homicides: ED use before victimization. Am J Emerg Med. 1999;17:689-691. 40. European Forum for Victim Services. The Social Rights of Victims of Crime. Brussels, Belgium: European Forum for Victim Services; 1998. 41. Shepherd JP, Morley R, Adshead G, et al. Ethical debate: should doctors be more proactive as advocates for victims of violence? BMJ. 1995;311:1617-1621. 42. Cretney A, Davis G, Clarkson C, et al. Criminal assault: the failure of the offence against society model. Br J Criminology. 1994;34:15-29. 43. Crisp R. Autonomy, welfare and the treatment of AIDS. In: Almond B, ed. AIDS: A Moral Issue. London, United Kingdom: Macmillan; 1990:68-91. 44. Sherman LW. Family based crime prevention. In: Sherman LW, Gottredson D, Mackenzie D, et al, eds. Preventing Crime: What Works, What Doesn’t, What’s Promising. Rockville, MD: National Criminal Justice Reference Service; 1997:4-9. NCJ 165366. 45. Dunford F. System-initiated warrants for suspects of misdemeanor domestic assault: a pilot study. Justice Q. 1990;7:631-653. 46. Pate AM, Hamilton EE. Formal and informal deterrents to domestic violence: the Dade County spouse assault experiment. Am Soc Rev. 1992;57:691-698. 47. Berk RA, Campbell A, Klap R, et al. The deterrence effect of arrest in incidents of domestic violence: a Bayesian analysis of four field experiments. Am Soc Rev. 1992;57:698-708.
20. Shepherd JP. Violence: the relation between seriousness of injury and outcome in the criminal justice system. J Accid Emerg Med. 1997;14:204-208.
48. Sherman LW, Smith DA. Crime, punishment and stake in conformity: legal and informal control of domestic violence. Am Soc Rev. 1992;28:183-206.
21. Shepherd JP, Lisles C. Towards multi-agency violence prevention and victim support: an investigation of police–accident and emergency service liaison. Br J Criminology. 1998;38:351-370.
49. Dunford FW, Huizinga D, Elliot D. The role of arrest in domestic assault. The Omaha Police Experiment. Criminology. 1992;28:183-206.
22. Langan PA, Farrington DP. Crime and Justice in the United States and in England and Wales 1981-96. Washington, DC: US Department of Justice; 1998. Publication NCJ 169284.
50. Ordre National des Medecins. Guide D’Exercise Professionel. Paris, France: Conseil National des Medecine; 1988.
23. Farrington DP, Langan PA. Changes in crime and punishment in England and America. Justice Q. 1992;9:5-46.
51. Shepherd JP. The casualty criminals. The London Times. July 16, 1996;Law section:31.
24. Farrington DP, Langan PA, Wikstrom PO. Changes in crime and punishment in America, England and Sweden between the 1980s and the 1990s. Stud Crime Prev. 1994;3:104-131. 25. Von Hirsch A, Bottoms AE, Burney E, et al. Criminal Deterrence and Sentence Severity: An Analysis of Recent Research. Oxford, United Kingdom: Hart Publishing; 1999.
52. Crime and Disorder Act 1998. Clause Five: Authorities Responsible for Strategies. London, United Kingdom: Her Majesty’s Stationery Office; 1998. 53. Shepherd JP. Cause for hope on crime statistics. The London Times. January 31, 2000:17. 54. Simmons S. Review of Crime Statistics: A Discussion Document. London, United Kingdom: Home Office; 2000.
26. Sivarajasingam V, Shepherd JP. Effectiveness of urban centre CCTV to prevent violence. J Accid Emerg Med. 1999;16:255-257. 27. Blumstein A, Cohen J, Nagin D, eds. Deterrence and Incapacitation. Washington, DC: National Academy of Sciences; 1978. 28. Reilly B, Witt R. Crime, deterrence and unemployment in England and Wales: an empirical analysis. Bull Econ Res. 1996;48:137-159. 29. Paternoster R. The deterrent effect of the perceived certainty of punishment. Justice Q. 1987;4:173-217. 30. Nagin D. Criminal deterrence research at the outset of the twenty-first century. Crime and Justice: A Review of Research. 1998;23:51-91. 31. Ross HL. Confronting Drunk Driving. New Haven, CT: Yale University Press; 1992. 32. Shepherd JP. Violent crime: using the NHS to unmask the scale of the problem. Policing Today. 1997;3:20-22. 33. Goodwin V, Shepherd JP. Evaluation of a questionnaire to facilitate the contribution of accident and emergency departments to Crime and Disorder Act local crime audits. J Accid Emerg Med. 2000;17:196-198. 34. Shepherd JP. The circumstances and prevention of bar-glass injury. Addiction. 1998;93:5-7. 35. Warburton AL, Shepherd JP. Effectiveness of toughened glassware in terms of reducing injury in bars: a randomised controlled trial. Injury Prev. 2000;6:36-40. 36. Blumstein A, Cohen J, Roth JA, et al. Criminal Careers and Career Criminals. Washington, DC: National Academy Press; 1988. 37. Farrington DP. Offending from 10 to 25 years of age. In: Van Dusen KT, Mednick SA, eds. Prospective Studies of Crime and Delinquency. Boston, MA: Kluwer Nijhoff; 1983. 38. Farrington DP, Wikstrom P-O H. Criminal careers in London and Stockholm. In: Weitekamp E, Kernen H-J, eds. Cross-national and Longitudinal Research on Human Development and Criminal Behaviour. Dordrecht, The Netherlands: Hewer; 1993.
OCTOBER 2001
38:4
ANNALS OF EMERGENCY MEDICINE
4 3 7
Emergency Medicine and Police Collaboration to Prevent Community Violence
From the Violence Research Group, University of Wales College of Medicine, Heath Park, Cardiff, United Kingdom. Received for publication February 3, 2000. Revisions received June 5, 2000, and September 28, 2000. Accepted for publication October 17, 2000. Address for reprints: Jonathan P. Shepherd, MSc, PhD, the Violence Research Group, University of Wales College of Medicine, Heath Park, Cardiff, CF14 4XY, United Kingdom; fax 00-44-2920-742442; E-mail [email protected]. Copyright © 2001 by the American College of Emergency Physicians. 0196-0644/2001/$35.00 + 0 47/1/114317 doi:10.1067/mem.2001.114317
Jonathan P. Shepherd, MSc, PhD
See related articles, p. 423 and p. 438. An important responsibility of emergency departments is the management of injuries sustained in assaults. Most assaults, including many causing serious injury, are not reported and not recorded by the police. This is important because police investigation and the conviction of offenders has a substantial deterrent effect and because information about the circumstances of intentional injury is key to prevention. Recent investigation of ED-police collaboration has shown that many of the injured, and ED staff want offenses to be reported but that there are attitudinal, logistic, and ethical-legal obstacles to achieving this. Organized joint efforts by emergency medicine personnel and police departments, on the basis of a sound legal and ethical framework to protect the rights of both victims and offenders, should deter more violent offenders and would-be violent offenders. They also provide the police with unique aggregate, nonconfidential information that is of substantial help in tackling violence. ED data can be used to measure and refine violence prevention initiatives and are being developed as the basis of a new, independent measure of police performance. Strategies, practical ideas to overcome obstacles, and directions for future research are suggested. [Shepherd JP. Emergency medicine and police collaboration to prevent community violence. Ann Emerg Med. October 2001;38:430-437.] INTRODUCTION
Emergency department personnel interface not only with medical services but also with local communities. They have extensive opportunities to work with local public services, of which police services are most important in maintaining law and order and preventing violence. In the context of violence as an important health issue, this article summarizes work that demonstrates the substan-
4 3 0
ANNALS OF EMERGENCY MEDICINE
38:4 OCTOBER 2001
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
tial extent to which violence, including violence that results in ED treatment, is not reported to the police; states why reporting is important; lists the obstacles to reporting; and identifies strategies and practical ideas about how obstacles might be overcome to the benefit of both communities and victims. It is concerned primarily with violence in which adults are injured. In setting out directions for future research, it identifies the need for evaluations of interventions that focus on both injury outcomes and process outcomes so that the benefits of ED-police collaboration can be maximized.
laws prevent violence and has been influential in improving criminal justice responses to handgun violence, alcohol-impaired driving, domestic violence, and child abuse. Although it might be assumed that criminal laws are built on good evidence of effectiveness, few have been quantitatively evaluated by using methods commonplace in medicine. The reasons for this include more reliance on qualitative approaches in legal and social sciences and reliance on incomplete and unreliable police crime statistics as measures of violent offenses.12 L O W P O L I C E R E P O R T I N G R AT E S F O R V I O L E N C E
I M P O R TA N C E O F V I O L E N C E A S A H E A LT H I S S U E
The importance of violence as a health issue is now well established. It is an important cause of death and physical and psychologic morbidity.1-3 Even in countries like the United Kingdom, which have strict weapon control laws, violence affects health substantially, particularly in terms of the facial deformity caused by blunt trauma.4,5 Leaving aside the injuries that violence produces, fear of crime can have a paralyzing effect on communities and can reduce quality of life for individual citizens.6 In turn, violence has been recognized as a public health problem, partly in the hope that, “It would be better to reduce violence through the positive aim of promoting health rather than through the negative aims of retribution, deterrence and incapacitation.”1,7 However, on their own, health, community safety, or criminal justice solutions to violence are not enough. There is now a consensus internationally that the key to community violence prevention is interagency coordination at the local level.8 Health services-criminal justice collaboration is important not only at the operational level but at the policy level as well. One of the most interesting recent developments in medicine is scrutiny of the effect of criminal laws by epidemiologists. It has been shown, for example, that accidental shooting deaths of children can be reduced by laws requiring gun owners to keep firearms locked away.9 An evaluation of state laws that restrict handgun purchase by persons with criminal records has demonstrated decreases in violent offenses committed by would-be purchasers who were denied handgun purchase.10 Targeted police work has been found to prevent alcohol-related violence in Australia by increasing deterrence.11 Studies like these suggest that law enforcement can produce changes in behavior. This health perspective of the criminal justice system is of great strategic importance. It provides a robust way of finding out whether criminal justice systems and criminal
OCTOBER 2001
38:4
ANNALS OF EMERGENCY MEDICINE
Another important reason to develop health servicescriminal justice partnerships is that EDs detect far more assaults in the community than are recorded by the police.12,13 Cross-referencing ED and police data concerning community violence has not been done systematically on a national basis, but studies in regional centers in the United Kingdom have shown that overall, only about 25% to 50% of offenses that led to ED treatment appeared in police records.12,13 This proportion is consistent with the findings of biennial British crime surveys, which compare householders’ accounts of crime with police records.14 In a UK study, assaults on men and assaults in bars and nightclubs and in the street that resulted in ED treatment were less likely to be recorded by the police than assaults on women and assaults in other locations.12 Differential rates of recording were also found according to time of day and day of week. This has obvious implications for the organization of community policing and law enforcement.12 In the United States, low rates of police recording-investigation of intimate partner violence have prompted legislation in some states to make reporting by health professionals compulsory.15 For firearm injury surveillance in the United States, although systems are being developed, systematic study of the extent to which there is correspondence between ED and police recording has yet to be done.16 A study of violence-related injuries treated in a sample of 31 US EDs in 1994 found that rates of injury inflicted by intimates (current or former spouse, boyfriend or girlfriend) were 4 times higher than rates from the National Crime Victimisation Survey.17 Because the purpose of this study was to estimate the overall burdens that violence imposes on EDs and not primarily to compare ED data with criminal justice data, the reasons for lack of police recording were not investigated. It was assumed, however, that such major differences could be ascribed to methodologic factors. Encouragingly for any future measurement of vio-
4 3 1
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
lence with ED records, however, there was close correspondence between overall rates of intentional injury when compared with rates from the 1994 National Hospital Ambulatory Medical Care Survey.18 Successive editions of the International Victimisation Survey have demonstrated that substantial numbers of violent offenses are not included in official crime statistics. They have also revealed that there are differences in rates of victimization between countries.19 For example, the 1996 survey showed that rates of victimization were highest in England and Wales, the United States, Sweden, and Finland and lowest in the Netherlands, Austria, and Northern Ireland. In the 11 industrialized countries that took part in the survey, the median reporting rate of violent offenses was 39% (range 18% to 51%).19 Although it might be thought that it is predominantly assaults causing minor injury that are not reported, 2 UK studies have found that police recording could not be predicted on the basis of injury severity scores and that many assaults causing severe injury are not reported.20,21 REASONS TO INCREASE REPORTING
If EDs and law enforcement worked together to enhance reporting of crimes, this could deter potential offenders, provide the police with information about violence that is not available from any other source, and help repair the wider damage done to victims and communities, potentially including the effects of restorative-reparative justice. There is now strong evidence that the police and the criminal justice process have a deterrent effect. International comparisons of police and crime survey measures of violence have demonstrated negative correlations between the chances of conviction and levels of violence.22-25 In 1981, the assault rate in England and Wales, according to victim surveys, was slightly higher than in the United States, but by 1995, it was more than double.22 The same trend is evident for other categories of violence. For murder, in 1985, rates in the United States were 8.7 times those of England and Wales, but by 1991, the excess risk had narrowed to 5.7 times. Over the same period, the proportion of assaults reported to the police in the United Kingdom did not change significantly, but improvements in recording rates for violence in the United States (63% recorded in 1981 to 93% recorded in 1995) were far greater than those noted in England and Wales (37% in 1981 to 46% in 1995). At the same time, there were obvious but opposite changes in rates of conviction for assault in the 2 countries.22 Between 1981 and 1995, conviction rates rose
4 3 2
steeply in the United States (from 0.16 to 0.44 per 1,000 population) but decreased markedly in England and Wales (from 1.12 to 0.61 per 1,000 population). A study of the effectiveness of urban closed-circuit television surveillance in Wales produced evidence consistent with this: only negative relationships were found between levels of community violence measured in EDs and levels according to police data.26 A recent authoritative review of research on criminal deterrence concluded that there had been negative correlations between the likelihood of conviction and crime rates studied over 10 years in England, the United States, and Sweden and over 15 years in England and the United States.25 This is consistent with previous research.27 Statistical associations between the severity of punishment and crime rates are much weaker than associations between certainty of punishment and crime rates and generally have not reached statistical significance.25 Although attempts have been made to control for influences on crime rates other than deterrence,28 the methods used have not been reliable enough to draw reliable conclusions.25 The principal source of quasiexperimental studies of deterrence is research on drunk driving.29,30 It has been concluded that, “There is considerable evidence that increasing the actual certainty of punishment for drunk driving in ways that also ensure adequate publicity of certainty of consequences can effect reductions in drunk driving.”31 Overall, the published evidence indicates that when would-be offenders perceive substantial risk of punishment, many are deterred from offending.25 This suggests that on the basis of a sound ethical framework, an important objective for primary and secondary injury prevention should be increasing the chances of punishment of violent offenders. This fits with the traditional public health strategy of targeting the agent of injury, in this case the offender or potential offender. Although treating violence as a public health issue is not new,1,7 formal collaboration between health and criminal justice services to maximize the chances of arresting and convicting violent offenders as a means of injury prevention has yet to be developed. An obvious application of this approach, given the low rates of recording by the police of offenses that result in ED treatment, is for EDs to facilitate and encourage increased rates of police reporting by or on behalf of the injured. The second main reason to improve reporting is that the circumstances of a great deal of community violence are not known to the police. Although it is often assumed
ANNALS OF EMERGENCY MEDICINE
38:4 OCTOBER 2001
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
that the police know about where injuries are sustained and which weapons have been used, this is often not the case.12,32 Just because police are often in EDs does not mean that they record and investigate all or even the majority of offenses. Information that can easily be obtained by ED personnel from victims33 is therefore often not available to the police and may be of substantial importance in targeting police resources to prevent violence or to arrest offenders. In cases with victims who do not want to report, this information can be provided by EDs in an aggregate, nonconfidential form. For example, because in the United Kingdom only about 1 in 9 assaults in licensed premises that result in ED treatment are recorded by the police,12 information from EDs can identify the bars and nightclubs that are the focus for violence. Similarly, EDs can provide unique information about the locations of work-, school-, and drug-related violence. In the United Kingdom, this ED perspective has even led to the identification and modification of a previously unrecognized weapon category, bar glassware, which is, in fact, the most frequently used weapon in the United Kingdom.34 Beer glasses in the United Kingdom are now manufactured almost exclusively of toughened glass. A recent randomized controlled trial shows that increasing the impact resistance of bar glasses reduces the risk of injury.35 If it had not been for ED-derived information, this advance would not have been made. It seems obvious that violence prevention should take account of risk of injury, as well as offenses reported to the police, but this does not seem to have been appreciated. Furthermore, a zero-tolerance policy demands that attention is paid to minor, as well as life-threatening, injuries. Increased reporting may not just be good for the community but also for patients because assaults are not usually isolated incidents in the lives of either victims or assailants. About half of assault victims treated in the ED have been assaulted before, and most violent offenses are committed by a small proportion of offenders.4,36,37 In a Philadelphia cohort study, it was found that 6% of male offenders (18% of all offenders) accounted for 52% of all arrests of juveniles. This was even more noticeable for violent offenses: 6% of chronic offenders accounted for 69% of all aggravated assaults, 71% of homicides, 73% of forcible rapes, and 82% of robberies.36,37 In London, similar results have been obtained: about 6% of male offenders (17% of all offenders) accounted for about 50% of all convictions.37 In Sweden, offending is even more concentrated.38 These data help to explain the high risk of repeat victimization. In a study in Kansas City of domestic homi-
OCTOBER 2001
38:4
ANNALS OF EMERGENCY MEDICINE
cide, a substantial proportion of victims were found to have attended local EDs for treatment of injuries sustained previously in cases of domestic violence.39 A collaborative approach also helps to safeguard the social rights of patients, particularly the right to privacy, and to prompt recognition and consideration of wider circumstances, expectations, and needs.40 Because having somewhere safe to live is essential to physical and psychologic well-being, rehousing on a permanent or temporary basis and increasing levels of security through, for example, lock or window replacement is important. Links between health and victim support services help to safeguard the dignity and social rehabilitation of the injured, as well as providing help and advice with claims for compensation. Several barriers to reporting have been identified: these may be categorized as attitudinal, logistic, ethical, and legal. A study of police-ED interaction carried out in 5 EDs in Wales, United Kingdom, showed that barriers to police reporting were, in decreasing order of importance, perceived to be the following: patient confidentiality, maintaining a neutral stance in relation to issues of blame, and protecting the patient.21 On the other hand, reasons for reporting, in decreasing order of importance, were perceived to be the following: reducing risk to others, severe injury, use of a weapon, and vulnerability of the injured. The skepticism with which the proposal that emergency physicians in the United Kingdom should be more proactive in reporting community violence to the police was met illustrates some of the attitudinal obstacles to the adoption of a more coordinated interagency approach to tackling violence.41 It was argued that violence prevention should not be medicalized because this would let official crime prevention agencies off the hook and divert some of the blame for violence to physicians. Working with the police to increase the chances of violent offenders being investigated was attacked as “unacceptable medical paternalism.”41 Trends toward partnership between public services to solve the problem of offending make these objections seem less appropriate now than previously.8 In contrast, in one study, half of emergency physicians, nurses, and clinical and managerial staff believed that they should have a role in victim protection and support, the detection of crime, and community crime prevention.21 Fewer obstacles have been identified in relation to attitudes of the police. In Wales, the police generally understood the ethical constraints on ED staff but wanted more violence to be reported. However, some police officers found it hard to come to terms with the reality of so much unreported violence and of the extra work this might pro-
4 3 3
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
vide.21 The need to prevent violence and protect patients and staff in EDs (eg, by ensuring that weapons are not brought into EDs) may divert police attention away from undetected community violence. The way the police respond to a report can influence whether an offense is recorded as well. The injured may be categorized as offenders, evidence may not be deemed to be sufficient, the victim may be unwilling to make a formal complaint, or the incident may be dismissed as a “family” affair.13 Victims’ attitudes are also important because the burden for supporting the prosecution process falls predominantly on them at every stage,13,42 including in the ED.21 The injured often do not go to the police because they cannot identify their assailant, they are marginalized as a result of alcohol or substance abuse, they are habituated to violence, they are unwilling to have their own conduct scrutinized, they are fearful of reprisals, or they have an ongoing relationship with their assailant.13 In this context, it is not surprising that even cases of serious gun violence can go unreported. However, many patients expect that once their medical needs are addressed, ED staff will report offenses on their behalf or at least rapidly summon police help to allow them to report.21 Victims and those who accompany them are not just concerned about physical injuries but also about the identification, arrest, and prosecution of the offender; prompt notification of the police; and prompt collection of evidence. Particularly at this early stage, patients want justice and look to ED staff to help them achieve this.21 In the ED, logistic obstacles to reporting include lack of facilities for patients to report, lack of a software facility to record basic information about the circumstances of violence, the often exclusively health agenda of EDs, and lack of links or interfaces between EDs and other community services, such as the police.21,33 Immediate availability of a police officer, police politeness, and preexisting professional relationships with police officers influence whether ED staff report offenses.21 Case studies illustrate these obstacles (Figure 1). In one study, the bureaucracy associated with reporting serious offences to the police led to informal police tip offs.21 Interestingly, in this study, individual clinicians had far more experience of contacting police officers than individual police officers had dealing with ED staff: relative inexperience of ED work may therefore be a logistic obstacle for the police. Neither group, however, knew about the other’s formal guidance or protocols. Conversations with ED receptionists were often perfunctory and limited to the recording of demographic
4 3 4
information, cause of injury, and previous ED visits. A potential logistic problem is the often protracted nature of ED treatment, which in the Wales study involved waits of up to 5 hours for the patients surveyed in waiting areas, in treatment cubicles, and on trolleys. These waits nearly always reflected how busy emergency physicians were, but they do provide clear opportunities and time for other staff, particularly receptionists and nurses, to institute discussion of reporting and victim support issues.21 Overall, victims of assault make up about 2% to 5% of all new ED attenders,4 a proportion that may make violence seem a relatively minor problem, particularly in countries in which firearm injuries are comparatively rare. Late at night, this proportion is very much higher, but at this time, senior staff who are in a position to develop violence-prevention strategies with the police may not be present.21 Although links between EDs and other medical services are well established, active partnerships between EDs and nonmedical public services are relatively undeveloped. The sequence of events exemplified in the case study from Wales (Figure 1) was noted time and time again during observation of ED staff-patient interaction: the injured and those who accompanied them repeatedly presented ED staff
Figure 1.
Case study. A dazed 20-year-old man with an obvious facial laceration reported to an ED receptionist after being assaulted in a bar. Although he was incapable of speech because of his injuries, his girlfriend kept making comments about wanting police involvement between trying to control her tears and physically supporting her boyfriend. She told the receptionist, “the police don’t know yet,” and when an ambulance driver passed the reception area, she asked, “Is that a policeman?” The receptionist ignored all these comments. Later, when the man was lying on an ED trolley, the girlfriend asked the male nurse who was swabbing his face, “Can you take photographs in here?” His reply was, “No ... you had better take them tomorrow when all the swelling comes out.” The girlfriend then immediately stated that after the attack in the pub, “I asked them to ring the police .... It was completely unprovoked.” Later, when the receptionist was asked by a research worker why she had not said anything about the police, she said that, “If she asked me, I would have told her ... we can’t do anything. They have to tell the police themselves.” When the research worker pointed out that people who had been attacked might not be aware of this, she agreed that, “If he hadn’t been assaulted before, he wouldn’t know the procedure.” Similarly, when the research worker asked the nurse why he had not responded more positively about taking photographs, he replied, “We haven’t got time to take photographs, and anyway, I wouldn’t want to go to court. It’s bad enough meeting patients in Tesco’s (a supermarket).” At the time, this male nurse appeared to have plenty of time: he was eating curry in the staff restroom. He stated that in relation to supporting victims, “we are taught not to be judgmental.”21
ANNALS OF EMERGENCY MEDICINE
38:4 OCTOBER 2001
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
with obvious cues to liaise with the police, but, usually under the guise of confidentiality, these were not acted on.21 Respecting the confidential nature of the physicianpatient relationship is of course of major importance, and apart from exceptional circumstances, it should be for patients to decide whether offences are reported. However, patients should be advised, in their medical interests and in the medical interests of others, to report offenses if there seems to be a risk of future harm. Importantly, the interests of patients and the wider community do not usually conflict.41 Conflict may arise, however, when the victim is a potential suspect. In the United Kingdom, physicians are advised to report offences, without the patients’ consent if necessary, in cases in which it appears that not doing so may hinder the detection, investigation, or prevention of a serious crime.21 Confidentiality can be an obstacle to responsible, ethical reporting if it is used by health professionals as a reason not to consider the needs of patients, communities, and the public services that serve them.21 Failure to consider these needs when treating a victim could, of course, lead to protection of offenders from prosecution. Of importance, however, is the possibility that unauthorized disclosure and a closer relationship between EDs and the criminal justice system might deter some people from seeking the medical help they need. Whether the knowledge that information may be disclosed actually stops people from seeking treatment is open to question, however.43 Few countries have laws designed to ensure the reporting of violent offenses by health professionals, although some US states have legislation to ensure that domestic violence is reported.15 The reasons for this include the sometimes fatal consequences of repeat violence and the threat that domestic violence poses to families and children. In the United Kingdom, reporting is left to the discretion of physicians. So far, however, on both sides of the Atlantic, there has been little or no requirement for ED staff to collect the information on which rational judgements about reporting can be made. In contrast to adult violence, in most countries, the investigation of child abuse has been the subject of legislation to ensure an appropriate multiagency approach. Surprisingly, although legislative steps that involve unauthorized disclosure of information have been taken to protect society from communicable disease, these have rarely been taken and enforced to protect society from violence, even though violence has become a leading cause of death, disfigurement, and disability.1-7 The lack of a legal framework to ensure that physicians make appropriate decisions about the reporting and
OCTOBER 2001
38:4
ANNALS OF EMERGENCY MEDICINE
investigation of adult violence may be an obstacle to protecting patients and communities. From an injury risk perspective, the focus in the United States on legislation concerning responses to domestic violence seems narrow. Very importantly, however, although legislation should protect the interests of citizens and communities, police responses to domestic violence can do more harm to victims than good.44-49 From an international standpoint, there are obviously potential abuses of closer collaboration between EDs and the police in countries with poor human rights records. For example, in France in 1944, attempts by occupying forces to oblige physicians to identify wounded members of the resistance prompted the President of the Ordre National des Medicines to remind all French physicians of their duty to preserve confidentiality. The text of his telegram still prefaces the “Guide d’Exercise Professionel.”50 In the United Kingdom, advocacy for a multiagency approach that breaks down barriers between services resulted in the Crime and Disorder Act of 1998.32,51,52 This act formally identified health authorities as partners with whom local government and the police must work to audit and tackle crime.52 I D E A S A B O U T H O W O B S TA C L E S C A N B E OVERCOME
Ideas about how these obstacles to reporting can be overcome are considered under the categories listed above. It is clear from the past research that some attitudes need to be changed. Education and training of ED personnel is needed to provide understanding of the wider needs of patients and the community in relation to deterrence, violence prevention, the causes of assault injury, the reporting of offenses, and the criminal justice process. Because the objective is ethical collaboration with the police, joint training seems important. An essential foundation of this training should be clear guidelines on circumstances in which reporting of offenses, without the patients’ consent if necessary, is justified. ED personnel (predominantly reception staff in the United Kingdom) who record details of offenses with suitably adapted software (Figure 2) also need training. Because this group has been shown to be in the best position to record information about the circumstances of violence and whether the assault has been reported to the police, training is necessary to enable them to do more than simply record administrative details.34 Minor changes to the organization of ED reception areas to facili-
4 3 5
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
tate the recording of what may be sensitive information may also be required. For example, patients and those who accompany them need to be able to have one-to-one conversations with receptionists away from lines or queues. Practical ways in which reporting in the ED can be facilitated and encouraged are listed in Figure 2. Because police crime statistics are known to be unreliable measures of community violence,12,19 assault injury rates are also a potentially invaluable measure of local, state, and national police performance and of the effectiveness of community violence prevention initiatives.53 The UK government has recently agreed that ED-derived data should be developed for this purpose.54 Further legislation in this area should be based on evidence of what works (see below) and should facilitate a coordinated approach to protect all adults and ensure access for victims to effective social and mental health interventions. It must also ensure that all serious violence is reported. DIRECTIONS FOR FUTURE RESEARCH
Although the strategies outlined in this article are rational in terms of assault injury prevention, it is vital that EDpolice partnerships on the basis of sound ethical principles are evaluated to find out whether they are effective. This is true both for existing and new partnerships. In
Figure 2.
Suggested intervention strategies.
particular, evaluations are needed of the use the police make of ED-derived reports of offenses and nonconfidential aggregate information about the circumstances of offenses and of joint interventions to target community violence. Evaluations should be focused not just on injury outcomes but also on process outcomes that local communities can adopt to ensure efficient and effective service coordination. In relation to the measurement of community violence and of police performance according to incidence and severity of assault injury, harmonization of police, and injury data sets is required. The effectiveness of laws mandating the reporting of domestic violence and, in the United Kingdom, of laws that include the health sector in community crime prevention also needs to be assessed. In summary, emergency physicians who treat victims of assaults should not assume that the police know about all or most violence. In all countries in which comparisons of police and crime survey data are possible, this has been shown not to be the case, even for serious offenses. Contacts with the ED are opportunities to improve reporting, deterrence, and violence prevention, as well as treatment and social care. While safeguarding the rights of both offenders and victims to treatment without fear of inappropriate disclosure, this is a rational step toward safer and more just communities in which the fear of crime is kept to a minimum. I thank Dr. Mike Maguire, School of Social and Administrative Studies, University of Cardiff, and Dr. David Farrington, Institute of Criminology, University of Cambridge, for their advice.
REFERENCES Strategies to increase rates of police reporting and investigation: • Free direct telephone lines from ED waiting areas to the police. • Incorporation of ED waiting areas into police patrols. • Asking all assault patients at the time of ED registration whether they would like assistance to report. • Reporting by ED staff on behalf of patients if they wish. • Mandatory reporting of serious violence. Strategies to target community policing more effectively: • Computerized collection of core data by admitting ED staff on the circumstances of community violence, particularly in the home, licensed premises, schools, and poor neighborhoods. • Regular disclosure to the police of nonconfidential aggregate information about assault time, weapons, assailants, and location. Strategies to measure community violence and police performance: • Development by the Centers for Disease Control and Prevention and National Institute of Justice of a core violence data set designed for ED use. • Inclusion of ED aggregate data in community audits of violence. • Publication of local, state, and national ED-derived information about community violence.
4 3 6
1.
Shepherd JP, Farrington DP. Assault as a public health problem. J R Soc Med. 1993;86:89-92.
2.
Rivara FP, Shepherd JP, Farrington DP. Gun laws. Lancet. 1993;342:809-810.
3. Joy D, Probert R, Bisson JI, et al. Post-traumatic stress reactions after injury. J Trauma. 2000;48:490-494. 4. Shepherd JP. Violent crime: an accident and emergency department perspective. Br J Criminology. 1990;30:289-305. 5. Bisson JI, Shepherd JP, Dhutia M. Psychological sequelae of facial trauma. J Trauma. 1997;43:496-500. 6.
Shepherd JP. Fear of crime. Lancet. 1990;335:796-797.
7. Moore MH. Public Health and criminal justice approaches to prevention. In: Tonry M, Farrington DP, eds. Building a Safer Society. Chicago, IL: University of Chicago Press; 1995:237-262. 8. Graham J, Ekblom P, Pease K, et al. Reducing Offending: An Assessment of Research Evidence on Ways of Dealing With Offending Behaviour. London, England: Home Office;1998. Research study 187. 9. Cummings P, Grossman DC, Rivara FP, et al. State gun safe storage laws and childhood mortality due to firearms. JAMA. 1997;278:1084-1086. 10. Wright M, Wintemute GJ, Rivara FP. Effectiveness of denial of handgun purchase to persons believed to be at high risk for firearm violence. Am J Public Health. 1999;89:88-90.
ANNALS OF EMERGENCY MEDICINE
38:4 OCTOBER 2001
POLICE COLLABORATION TO PREVENT COMMUNITY VIOLENCE Shepherd
11. Homel R, Hauritz M, Wortley A, et al. Preventing alcohol-related crime through community action: the surfers’ paradise. In: Homel R, ed. Policing for Prevention. Crime Prevention Studies 7. Monsey, New York, NY: Criminal Justice Press; 1997. 12. Shepherd JP, Shapland M, Scully C. Recording of violent offences by the police: an accident and emergency department perspective. Med Sci Law. 1989;29:251-257. 13. Clarkson C, Cretney A, Davis G, et al. Assaults: the relationship between seriousness, criminalisation and punishment. Criminal Law Rev. 1994;Jan:4-21. 14. Mirrlees-Black C, Budd T, Partridge S. British Crime Survey: England and Wales 1998. London, England: Home Office; 1998. 15. Cole TB, Flanigan A. What can we do about violence? JAMA. 1999;282:481-483. 16. Kellermann AL, Bartolomeos KK. Firearm injury surveillance at the local level: from data to action. Am J Prev Med. 1998;15:109-112. 17. Rand MR, Strom K. Violence-related Injuries Treated in Hospital Emergency Departments. Washington, DC: US Department of Justice; 1997. Publication NCJ-156921. 18. National Hospital Ambulatory Medical Care Survey (NHAMCS). Atlanta, GA: Centers for Disease Control; 1994. 19. Mayhew P, Van Dijk JJM. Criminal Victimisation in Eleven Industrialised Countries. Key Findings from the 1996 International Crime Victims Survey. Amsterdam, The Netherlands: WODC Ministry of Justice; 1997.
39. Wadman MC, Muelleman RL. Domestic violence homicides: ED use before victimization. Am J Emerg Med. 1999;17:689-691. 40. European Forum for Victim Services. The Social Rights of Victims of Crime. Brussels, Belgium: European Forum for Victim Services; 1998. 41. Shepherd JP, Morley R, Adshead G, et al. Ethical debate: should doctors be more proactive as advocates for victims of violence? BMJ. 1995;311:1617-1621. 42. Cretney A, Davis G, Clarkson C, et al. Criminal assault: the failure of the offence against society model. Br J Criminology. 1994;34:15-29. 43. Crisp R. Autonomy, welfare and the treatment of AIDS. In: Almond B, ed. AIDS: A Moral Issue. London, United Kingdom: Macmillan; 1990:68-91. 44. Sherman LW. Family based crime prevention. In: Sherman LW, Gottredson D, Mackenzie D, et al, eds. Preventing Crime: What Works, What Doesn’t, What’s Promising. Rockville, MD: National Criminal Justice Reference Service; 1997:4-9. NCJ 165366. 45. Dunford F. System-initiated warrants for suspects of misdemeanor domestic assault: a pilot study. Justice Q. 1990;7:631-653. 46. Pate AM, Hamilton EE. Formal and informal deterrents to domestic violence: the Dade County spouse assault experiment. Am Soc Rev. 1992;57:691-698. 47. Berk RA, Campbell A, Klap R, et al. The deterrence effect of arrest in incidents of domestic violence: a Bayesian analysis of four field experiments. Am Soc Rev. 1992;57:698-708.
20. Shepherd JP. Violence: the relation between seriousness of injury and outcome in the criminal justice system. J Accid Emerg Med. 1997;14:204-208.
48. Sherman LW, Smith DA. Crime, punishment and stake in conformity: legal and informal control of domestic violence. Am Soc Rev. 1992;28:183-206.
21. Shepherd JP, Lisles C. Towards multi-agency violence prevention and victim support: an investigation of police–accident and emergency service liaison. Br J Criminology. 1998;38:351-370.
49. Dunford FW, Huizinga D, Elliot D. The role of arrest in domestic assault. The Omaha Police Experiment. Criminology. 1992;28:183-206.
22. Langan PA, Farrington DP. Crime and Justice in the United States and in England and Wales 1981-96. Washington, DC: US Department of Justice; 1998. Publication NCJ 169284.
50. Ordre National des Medecins. Guide D’Exercise Professionel. Paris, France: Conseil National des Medecine; 1988.
23. Farrington DP, Langan PA. Changes in crime and punishment in England and America. Justice Q. 1992;9:5-46.
51. Shepherd JP. The casualty criminals. The London Times. July 16, 1996;Law section:31.
24. Farrington DP, Langan PA, Wikstrom PO. Changes in crime and punishment in America, England and Sweden between the 1980s and the 1990s. Stud Crime Prev. 1994;3:104-131. 25. Von Hirsch A, Bottoms AE, Burney E, et al. Criminal Deterrence and Sentence Severity: An Analysis of Recent Research. Oxford, United Kingdom: Hart Publishing; 1999.
52. Crime and Disorder Act 1998. Clause Five: Authorities Responsible for Strategies. London, United Kingdom: Her Majesty’s Stationery Office; 1998. 53. Shepherd JP. Cause for hope on crime statistics. The London Times. January 31, 2000:17. 54. Simmons S. Review of Crime Statistics: A Discussion Document. London, United Kingdom: Home Office; 2000.
26. Sivarajasingam V, Shepherd JP. Effectiveness of urban centre CCTV to prevent violence. J Accid Emerg Med. 1999;16:255-257. 27. Blumstein A, Cohen J, Nagin D, eds. Deterrence and Incapacitation. Washington, DC: National Academy of Sciences; 1978. 28. Reilly B, Witt R. Crime, deterrence and unemployment in England and Wales: an empirical analysis. Bull Econ Res. 1996;48:137-159. 29. Paternoster R. The deterrent effect of the perceived certainty of punishment. Justice Q. 1987;4:173-217. 30. Nagin D. Criminal deterrence research at the outset of the twenty-first century. Crime and Justice: A Review of Research. 1998;23:51-91. 31. Ross HL. Confronting Drunk Driving. New Haven, CT: Yale University Press; 1992. 32. Shepherd JP. Violent crime: using the NHS to unmask the scale of the problem. Policing Today. 1997;3:20-22. 33. Goodwin V, Shepherd JP. Evaluation of a questionnaire to facilitate the contribution of accident and emergency departments to Crime and Disorder Act local crime audits. J Accid Emerg Med. 2000;17:196-198. 34. Shepherd JP. The circumstances and prevention of bar-glass injury. Addiction. 1998;93:5-7. 35. Warburton AL, Shepherd JP. Effectiveness of toughened glassware in terms of reducing injury in bars: a randomised controlled trial. Injury Prev. 2000;6:36-40. 36. Blumstein A, Cohen J, Roth JA, et al. Criminal Careers and Career Criminals. Washington, DC: National Academy Press; 1988. 37. Farrington DP. Offending from 10 to 25 years of age. In: Van Dusen KT, Mednick SA, eds. Prospective Studies of Crime and Delinquency. Boston, MA: Kluwer Nijhoff; 1983. 38. Farrington DP, Wikstrom P-O H. Criminal careers in London and Stockholm. In: Weitekamp E, Kernen H-J, eds. Cross-national and Longitudinal Research on Human Development and Criminal Behaviour. Dordrecht, The Netherlands: Hewer; 1993.
OCTOBER 2001
38:4
ANNALS OF EMERGENCY MEDICINE
4 3 7