The Content of Management of Violence Policy Documents

HEALTH AND NURSING POLICY ISSUES The content of management of violence policy documents in United Kingdom acute inpatient mental health services James Noak Steve Wright Jane Sayer RN RMN MSc Robert Baxter Research Fellow, Health Service Research Department, Institute of Psychiatry, London, UK BSc DipPsychol PGDip MSc Research Worker, Health Service Research Department, Institute of Psychiatry, London, UK BA MSc RN Deputy Director of Nursing, South London and Maudsley Trust, London, UK Ann-Marie Parr Richard Gray BSc Research Worker, Health Service Research Department, Institute of Psychiatry, London, UK BSc PhD RN MRC Research Fellow, Health Service Research Department, Institute of Psychiatry, London, UK Dylan Southern MBA RMN RGN Executive Director of Care Services, Ashworth Hospital, Merseyside, UK and Kevin Gournay CBE PhD CPsychol RN Professor of Psychiatric Nursing, Health Service Research Department, Institute of Psychiatry, London, UK Submitted for publication 8 May 2001 Accepted for publication 21 November 2001 N O A K J ., W R I G H T S ., S A Y E R J ., P A R R A -M ., G R A Y R ., S O U T H E R N D . & M GOURNAY K. (2002) Correspondence: James Noak, Health Service Research Department, Institute of Psychiatry, De Crespigny Park, Camberwell, London SE5 8AF, UK. E-mail: j.noak@jop.kcl.ac.uk Journal of Advanced Nursing 37(4), 394±401 The content of management of violence policy documents in United Kingdom acute inpatient mental health services Aim of the study. The aim of the study was to examine the content of Trust policies concerning the prevention and management of violence in acute in-patient settings in order to establish their usefulness as guidance for staff in this dif®cult, complex, and controversial aspect of inpatient psychiatric care. Background. Violence is a commonly encountered problem in inpatient psychiatric settings. There are legal requirements for workplaces in general and mental health care facilities in particular to develop safe systems of work based upon the ®ndings of assessments of this risk. Policies have a key role to play in making explicit the responsibilities of both employer and employees, and specifying standards of acceptable practice. Design. A cross-sectional survey methodology was used, which entailed examination of the content of management of violence policies that had been forwarded to the authors from 40 Trusts providing acute inpatient psychiatric care throughout England, Scotland, Wales, and Northern Ireland. Findings. Policies were found to vary widely in their content, and serious shortcomings were noted in the extent to which policies included information Ó 2002 Blackwell Science Ltd 394 Health and nursing policy issues Management of violence policy review regarding their status and review, advice on the prevention of violence, the management of violent incidents, and postincident action. Conclusions. Further research is needed to tease out the extent to policies which are lacking in content, re¯ect shortcomings in the organizational approach to the prevention and management of violence by Trusts, and the extent to which such shortcomings result in harm being suffered by staff and/or patients. An alternative format for the presentation of management of violence policies is discussed, and items that should be included in inpatient units' management of violence policies are suggested. Keywords: violence, mental health policy, policy formulation, inpatient care Introduction Violence is a key issue affecting public service workers. Front line workers such as police of®cers, ®re ®ghters, and health care workers are consistently shown to be among those most at risk. Health care workers are often the target of violence (Rippon 2000), and the environments in which they work are becoming increasingly dangerous places to work (Rosenthal et al. 1992, Dillner 1994), with nurses appearing to be most at risk of becoming a victim of violence at work. Indeed, the Health & Safety Executive (HSE) (1997a, 1997b) suggested that the risk of violence faced by nurses could be as much as ®ve times greater than that faced by the general population. While there is considerable variation in the number of incidents reported, the average number of incidents in mental health and learning disability settings was over three times greater than the average across Trusts in general. Violence against mental health care staff can come from several sources such as visitors, patients, intruders, and fellow workers. Apart from the number of assaults against staff, evidence also suggests that people who use mental health services are often the victims of violence themselves (e.g. Estroff & Zimmer 1994) and overall, acute inpatient mental health units are perceived as dangerous environments for patients and staff alike (Dillner 1994, Gournay et al. 1998). Violence is de®ned by Steinmetz and Lystadad (1986, p. 52) as `an act carried out with the intention or perception as having the intention, of physically hurting another person'. This de®nition includes all incidents from minor assault to premeditated murder. The effects of physical assault are obvious and wide-ranging, and it should be acknowledged that while physical violence can clearly result in physical injury, it frequently has an emotional impact for the victim, and possibly for witnesses. Perpetrators of violence can also suffer psychological and emotional distress, and verbal aggression in the form of abuse and threats may also result in considerable emotional damage (see Adams & Whittington 1995). The situation is perhaps more serious than it appears because the majority of assaults which occur in mental health care settings go unreported (see Crowner et al. 1994, Thomas et al. 1995). Violence by inpatients and its management, has also been identi®ed as a major source of stress among nursing staff in inpatient settings (see Prosser et al. 1997, Reid et al. 1999). Nursing staff who have actually been assaulted have been found to suffer from a range of anxiety-related symptoms, which inevitably have implications for their ability to work effectively (Whittington & Wykes 1992), and can also result in higher levels of sickness and absenteeism than are observed among their nonassaulted colleagues (Gournay et al. 2000). The possibility that assaultative patients receive substandard care because mutual hostility between patient and staff fuels the breakdown of the therapeutic alliance, has also been discussed (Lipsedge 1994, Watts & Morgan 1994). The problem of violence in health care settings must be addressed at a variety of levels, leading to the need for a consistent, co-ordinated organizational response. Violence is therefore presents as an organizational issue. Policies are the cornerstone of an organizational approach to any problem because they make explicit the responsibilities of both employer and employees, and specify standards of acceptable practice. The Management of Health and Safety Regulations (Health and Safety Commission 1992) require employers to conduct assessments of likely sources of risk to employees and others present in the workplace, and to implement a safe system of work. Employers have a legal responsibility to take steps to manage the risks highlighted by risk assessment, to allocate appropriate people to assist in the formulation and implementation of the necessary measures, to establish emergency procedures, and to ensure that employees have adequate training and are suf®ciently 395 Ó 2002 Blackwell Science Ltd, Journal of Advanced Nursing, 37(4), 394±401 J. Noak et al. competent in their work to avoid risks. In mental health care settings, the Mental Health Act (1983) Code of Practice (Department of Health & Welsh Of®ce 1999) states that all providers of in-patient psychiatric care should have clear policies on the use of restraint in the management of violence. Policies are credible and useful to the extent that they provide clear guidance on practice. In the research that underpinned the Standing Nursing Midwifery Advisory Committee report `Addressing Acute Concerns' (Department of Health 1999) a team from the Institute of Psychiatry reviewed the content of management of violence policies in 33 Trusts operating regional secure and psychiatric intensive care units (Wright et al. 2000). This research found that policies on the management of violence often lacked guidance on important areas. The present study builds on the SNAMAC report and focuses on a survey of Trusts across the United Kingdom (UK) which provide acute inpatient mental health services. The study Method Design A cross-sectional postal survey methodology was used. Sample The Institute of Health Service Management Yearbook was used to identify all National Health Service (NHS) Trusts in the UK who were listed as currently providing inpatient psychiatric care, and 188 Trusts were initially identi®ed for possible inclusion in the study (155 in England, 18 in Scotland, 10 in Wales and ®ve in Northern Ireland). However, changes in NHS con®guration meant that several trusts no longer provided mental health care when the survey was about to be launched, and these were therefore removed from the sampling frame. The 33 Trusts whose policies had previously been reviewed by Wright et al. (2000) were also excluded. Finally a sample of 40 Trusts (in which all four countries in the UK were represented) was randomly selected from the remainder. Procedure All 40 trusts which were approached agreed to participate. A letter was sent to Directors of Nursing (or equivalent) explaining the basis of the research and asking for a copy of any policies that related to the management of violence within their Trust. If there was no response after one calendar month the Head of Nursing of each Trust was contacted by 396 telephone and was asked if the trust had received the letter. The consent of each Head of Nursing for participation in the study was requested, and any remaining queries about the study were resolved. Non-responders were contacted at twoweekly intervals for 6 weeks. When the policies arrived and were collated, they were each rated independently by two experienced mental health nurses (a Deputy Director of Nursing and a Research Fellow), using the template used by Wright et al. (2000). This template is based on recommended features of management of violence policies in social services and probation settings described by Kedward (1990). It also contains additional items concerning the development and rati®cation of the policy, and items pertaining to good practice and training in the prevention and management of violence derived from the Mental Health Act (1983) Code of Practice (Department of Health & Welsh Of®ce 1999) and which are more relevant to in-patient psychiatric settings. Given the concerns expressed by mental health professionals and the Police Complaints Authority regarding the possible use of CS incapacitant spray by police in the management of violent patients (Bell & Thomas 1998, Police Complaints Authority 2000), and given the recently reported risk of burns to skin exposed to CS spray (Southward 2001), policies were also examined for guidelines concerning the use of CS spray on hospital premises and regarding the management of patients who had been exposed to it. Data analysis Out of 1517 pairs of ratings, inter-rater agreement was observed in over 99% of cases. Where the independent raters disagreed on whether an item was included in a policy, the document was jointly examined and a consensus was arrived at as to whether or not the item in question was present, and a rating was made accordingly. Results For convenience, the results may be considered under the summary headings of organizational issues, training and education, guidance on the management of violent situations, and post incident responses, and are presented below. The items that were included in the template, and the frequency with which each item was rated as present within each policy, are presented in Table 1. Organizational issues Three Trusts (8%) stated that they had no current policy on the management of violence. Most policies (76%) gave a Ó 2002 Blackwell Science Ltd, Journal of Advanced Nursing, 37(4), 394±401 Health and nursing policy issues Table 1 Policy items included in the template, and the frequency with which each item was rated as included in the policies (n ˆ 40)* Management of violence policy review Item Statement of responsibility on the part of the authority Emphasis on physical restraint as measure of last resort Commitment to appropriate training De®nition of violence Is the policy dated? Potential causes of violence mentioned Statement of the aims of the policy Emphasis on use of minimum or reasonable force Purpose of restraint Mention of need to call for help Methods of coping: (a) de-escalation Expectations and responsibilities of staff Post-incident analysis and support for staff Patient complaint procedure outlined Reporting of incident for audit mentioned Reporting of incident in patient's case notes mentioned Preventative measures emphasized Need for one staff member to take control of the incident mentioned Post incident review of care plan Circumstances under which police assistance should be sought described Information regarding warning signs of imminent violence Mention of unacceptable methods of restraint Mention of need for refresher training Occupational health/other staff welfare provision mentioned Information about support for legal help/compensation Need to maintain communication with the patient emphasized Methods of coping: (b) breakaways Acceptable reasons for restraint given Methods of coping: (c) physical restraint Identi®cation of those responsible for ratifying, monitoring and evaluating the policy Is the policy up to date (i.e. dated within the review period)? Is there reference to when the policy should be reviewed? Other patients to leave scene of the incident Intervals for refresher training speci®ed Mention of need for visual check for weapons Reference to MHA Code of Practice Some account of the incidence of violence/threats Post-incident analysis and support for other patients Post-incident analysis and support for assaultative patient Any mention of use of CS spray by police in clinical areas Any advice on care of patients exposed to CS spray *Three Trusts had no policy. Inclusion (%) 79 79 76 76 70 67 67 67 67 67 64 64 64 61 61 55 55 55 52 52 48 48 45 45 45 45 42 42 41 39 39 36 33 30 30 27 24 15 12 0 0 de®nition of violence which referred to both psychological and physical harm, 79% of policies included a statement of the employing organization's responsibility on the part of the Trust, and 67% stated clear aims for their policy. Thirty-nine per cent of the policies identi®ed who was responsible for ratifying, monitoring, and evaluating the policy. Seventy percent of the policies had a implementation date and 36% included a review date. Thirty-nine percent of policies with a review date were de®nitely up-to date (i.e. with an implementation date within the stated review period). Training and education Most policies (76%) expressed a commitment to training, but less than half (45%) mentioned the need for refresher training, and fewer still (30%) speci®ed how frequently refresher training should be taken. Two-thirds of the policies mentioned potential causes of violence, just over half (55%) gave information on the available preventative measures to minimize violence, and fewer than half (48%) discussed common behavioural warning signs of imminent violence. 397 Ó 2002 Blackwell Science Ltd, Journal of Advanced Nursing, 37(4), 394±401 J. Noak et al. Practical management Regarding the management of potentially or actually violent situations, de-escalation strategies were mentioned in 64% of policies, breakaway techniques were referred to in 42%, and physical restraint was mentioned in 41% of policies. Regarding management on the scene, 55% of policies mentioned the need for one member of staff to take control of the situation. The need for communication with patients was mentioned in 55% of policies, and a third pointed out the need to remove other patients from the scene. Thirty per cent of policies mentioned the need to check visually for weapons. The circumstances of when the police should be called was included in 52% of policies, but no policies gave advice regarding the use of CS incapacitant spray by police in clinical areas, or advice on the care of patients or staff who had been exposed to CS spray. Post incident analysis and support The need for postincident analysis and support for staff is included in the majority of polices (64%), but the need for such support for other patients who witnessed the incident and for the assaultative patient was present infrequently (in 15% and 12% of policies, respectively). Just over half (52%) of the policies suggested that there might be a need for a postincident review of the assaultative patient's care plan. Patients' complaints procedures were included in 61% of the policies, and information on legal help, compensation and staff welfare was included in 45%. Discussion The 100% response rate from the 40 Trusts was encouraging and may compensate to some degree for the comparatively small sample size. Overall, the study's ®ndings largely con®rm those reported in other studies (e.g. Oakley 1998, Wright et al. 2000). It is tempting to consider in detail the de®cits uncovered in Trust policies and to explore their possible impact upon clinical practice and clinical and corporate governance, but this would be beyond the scope of this paper. Oakley (1998) stated that 7% of NHS trusts had no policy on violence, and a similar ®gure was found in this study. This absence of policy should be great cause for concern for the organizations involved and for clinicians in practice. It ¯ies in the face of current guidelines, it leaves clinicians vulnerable to confusion, and renders organizations vulnerable to legal action from staff, patients and statutory bodies. As discussed above, the Health and Safety Commission (1992) makes 398 explicit the legal duty of employers to take reasonable steps to ensure the health and safety of employees and others in the workplace, and the Mental Health Act (1983) Code of Practice states the need for clear policies on the use of restraint in the management of violence. The fact that less than half of the policies reviewed were clearly dated within their own stated review period is a further poor re¯ection upon the organizational structures that are responsible for policy administration. The ®nding that comparatively few of the policies reviewed emphasized the prevention of violence or gave advice concerning preventative strategies, or mentioned behavioural cues associated with imminent violence, also runs counter to the stress placed on prevention in the Code of Practice. The lack of advice regarding management strategies also gives cause for concern. Few policies speci®ed the need for one staff member to assume control of the incident, for a visual check for weapons to be required, for other patients to be made to leave the vicinity, or for the need to communicate with the assaultative patient so as to give the opportunity for voluntary co-operation. Few policies gave helpful advice regarding the updating in practical skills in the management of violence, and only just over half gave advice on the circumstances under which to call for police assistance. Given the current concerns surrounding the possible use of CS spray by police when assisting in the management of violent incidents involving mental health service users, both within and outside of clinical care environments, it is worrying to ®nd that there was no advice regarding this present in any of the documents that were examined. This is despite the fact that CS spray has been carried operationally by police forces on a nationwide scale since March 1996, and Bell and Thomas (1998) reported that 31 out of 108 Trusts (29%) that they surveyed reported that patients had been admitted suffering from the effects of CS spray exposure, and that such sprays had been used on admission wards. The present study ®ndings suggest that the situation regarding guidelines on the use of CS spray on hospital premises remains essentially unchanged. Bell and Thomas (1998) reported that only one Trust in their survey had produced guidelines regarding the use of CS spray on NHS premises, and only two reported having guidelines on the handling of patients who had been exposed. Several of the concerns highlighted by Bell and Thomas are also mentioned in the Police Complaints Authority (PCA) report into the safety and effectiveness of CS spray (PCA 2000). The PCA report also mentions concerns relating to interactions between CS spray and antipsychotic medication, and highlights the effective criminalization of mental disorder that may be perceived in its use on people with such Ó 2002 Blackwell Science Ltd, Journal of Advanced Nursing, 37(4), 394±401 Health and nursing policy issues Management of violence policy review conditions, and advises consultation with relatives and professionals before its use. This lack of guidance is clearly in need of urgent remedy. It is also disconcerting to ®nd that postincident support for the assaultative patient and other patients is mentioned in only a minority of policies. While patients' complaints procedures were included more frequently, it is disappointing that the mechanisms by which complaints should be registered were not detailed more often. The management of violent patients remains a controversial issue, and given the widespread concerns regarding the possibility of abuse in the course of physical restraint (see Allen & Harris 2000). The mandatory inclusion in management of violence policies of how complaints may be made by patients or their advocates should be seriously considered. This would register a commitment to the preservation of patients' human rights on the part of organizations, and may encourage scrupulous attention to good practice in practitioners. By the same token, the fact that information on legal help, compensation and staff welfare resources available to victimized staff was included in less than half of the policies reviewed implies a lack of concern for staff welfare. This does not assist in ensuring consistent support for victims, and might actually lead to secondary victimization. organizational-level responses to the issue as operationalized in Trust policies. The consent of Trusts was sought on this basis, and so it was not possible to ask for performancerelated data. Furthermore, given the complexity and size that such a task would have presented, this would have been beyond the limited scope of this part of the overall initiative. The relationship between policy content and practice It is acknowledged that the issue of the relationship between policy content and operational performance in this area is very important. Dowson et al. (1999) found several shortcomings in performance where provision that should be addressed by policies was de®cient. Dowson et al. (1999) conducted a survey of management of violent incidents in a total of 15 sites containing general adult psychiatric inpatient units (GPUs) and learning disability inpatient units (LDUs) in 10 NHS Trusts in the Anglia Region. While a written policy for the management of violent incidents was present on the unit where 98% of incidents occurred, 41% of the incidents where physical restraint was used (and where information could be obtained on all staff engaged in the procedure) involved some staff who were not acquainted with Trust policy. Furthermore, only 27% of such incidents recorded in GPUs and none of those in LDUs involved staff who had all received training or training updates in the 12 months prior to the incident. While most of these incidents were resolved without injury to either patients or staff, 61% of those incidents where an injury to staff occurred and over half of those where injury to a patient occurred failed to meet the standard of staff training detailed above. The standard of reporting for 97% of the incidents recorded in the GDUs and in 85% of the LDUs was deemed to be insuf®cient, and only four of the sites involved had a written policy related to the management of violence which included procedures related to victim support. Of all the staff interviewed who had been physically assaulted, 16% of those employed in GPUs and 24% of those employed in LDUs considered that postincident support had been inadequate. While it is impossible to say whether, on the basis of the data presented, these shortcomings occurred as a direct result of this lack of provision, or whether these areas were covered in the policies of the units concerned, these questions are raised nevertheless, and merit fuller research attention in future. Study limitations The survey's results show an apparently wide variability in the content of management of violence policies in acute mental health services. However, this variation needs to be considered in the light of the study's methodological limitations. While the sample was considered to be representative of mental health Trusts in the UK, it is modest in size, and so its results should be considered in this light. It is also possible that aspects of the prevention and management of violence may be contained within policies other than those explicitly concerned with the prevention and management of violence, which were not forwarded to the research team. However, this in itself raises the question of staff awareness of which policy to consult for guidance, and their accessibility. Another limitation of the study is that it is essentially descriptive, and does not attempt to explore differences in performance in the safe and professionally acceptable prevention and management of violence, and other characteristics between those Trusts whose policies were comprehensive, and those whose were not. This was because the survey was conducted as part of a United Kingdom Central Council for Nursing, Midwifery and Health (UKCC) initiative concerning the prevention and management of violence in acute inpatient settings, and sought to provide an overview of Conclusion: points for action It is plainly unrealistic to expect any policy to provide exhaustive detail as to how staff should act in all situations. However, this study highlighted some clear omissions of key 399 Ó 2002 Blackwell Science Ltd, Journal of Advanced Nursing, 37(4), 394±401 J. Noak et al. areas in the policies examined. If policies are meant to assist staff in safeguarding service users and to provide staff with guidance regarding safe and acceptable practice, then they must be clear and unambiguous, and should address the contingencies that can reasonably be expected to occur. This implies that they should be based on the best available evidence. Policies that affect clinical practice should also be clearly in line with clinical governance. The form in which policies are presented may also bear some reconsideration. Large, comprehensive written polices may actually be unwieldy or confusing, and may discourage staff from consulting them. It might be more useful to present policies in diagrammatic form (as ¯ow charts, for example), which may be simpler and more readily understandable. If the conventional format of policy documents is maintained, then the policies should at least be written in a way that is understandable to all staff, and therefore care should be taken to use effective and simple language. It is also vital that staff are well trained and practised in the recommended methods of preventing and managing violence, which, given the circumstances, ought to be regarded as a core skill for `front line' staff. Besides highlighting the need for further research examining how effective comprehensive policies are in the prevention and management of violence, two avenues for further action are suggested by the survey. First, given the increasingly litigious climate in which health care services operate, health care providers may ®nd it prudent to examine their current policies from a legally defensive standpoint in the light of the survey's ®ndings. Second, attention should be given to ensuring that policies are as comprehensive as possible. The development of mandatory minimum standards for policy content would be of value, which should be informed by reference to guidance and good practice initiatives, that has already been provided in the Mental Health Act (1983) Code of Practice, or has been established by bodies such as the Royal College of Psychiatrists (1998), and which is also forthcoming from the UKCC initiative on the prevention and therapeutic management of violence and aggression. In reviewing these polices and in examining the literature the research team felt some further areas needed inclusion included; gender issues, issues around sensory handicaps for patients and staff, the area of pregnancy for patients and staff as well as seclusion and the processes surrounding rapid tranquillization, including the nurses' and physicians' roles and responsibilities in rapid tranquillization. A recommended list of items for inclusion in management of violence policies which draws on these guidelines and on other issues arising from them, and from other research (Lee et al. 2001) is suggested in the Appendix. 400 Acknowledgements This review of management of violence policies was part of a study commissioned by the United Kingdom Central Council for Nursing, Midwifery and Health visiting. The full study was completed by the members of staff from Health Service Research Department at the Institute of Psychiatry. Comments made by the authors do not necessarily represent the views of the UKCC or the Institute of Psychiatry. References Adams J. & Whittington R. (1995) Verbal aggression to psychiatric staff: traumatic stressor or part of the job? Psychiatric Care 2, 171±174. Allen D. & Harris J. (2000) Abuse by any other name: a critique of some current approaches to behaviour management. Mental Health Care 3, 188±189. Bell F. & Thomas B. (1998) Police use of CS spray: implications for NHS mental health services. Mental Health Care 1, 402±404. Crowner M.L., Peric G., Stepcic F. & Van Oss E. (1994) A comparison of videocameras and of®cial incident reports in detecting inpatient assaults. Hospital and Community Psychiatry 45, 1144±1145. Department of Health & Welsh Of®ce (1999) Mental Health Act 1983 Code of Practice. HMSO, London. Department of Health. (1999) Standing Nursing Midwifery Advisory Committee Addressing Acute Concerns. Department of Health, London. Dillner L. (1994) Psychiatric services are dangerous places to work. British Medical Journal 309, 77. Dowson J.H., Butler J. & Williams O. (1999) Management of psychiatric in-patient violence in the Anglia region. Psychiatric Bulletin 23, 486±489. Estroff S.E. & Zimmer C. (1994) Social networks, social support, and violence among persons with severe, persistent mental illness. In Violence and Mental Disorder: Developments in Risk Assessment (Monahan J. & Steadman H.J. eds) Chicago University Press, Chicago. Gournay K., Carson J. & Spence A. (2000) Special Hospitals' Stress Survey (unpublished report). Institute of Psychiatry, London. Gournay K., Ward M., Thornicroft G. & Wright S. (1998) Crisis in the capital: in-patient care in inner London. Mental Health Practice 1, 10±18. Health and Safety Commission (1992) The Management of Health and Safety at Work: Approved Code of Practice. HMSO, London. Health and Safety Executive (1997a) UK Household Survey. Health and Safety Executive Books, London. Health and Safety Executive (1997b) Violence and Aggression to Staff in the Health Service: Guidance on Assessment and Management. Health Service Executive Books, London. Kedward C. (1990) Current practice and policy with regard to violence in social services departments in the UK the University of Sussex Study. In Violence Against Social Workers: Implications for Practice (Norris D. ed.) Jessica Kingsley, London. Lee S., Wright S., Sayer J., Parr A.-M., Gray R. & Gournay K. (2001) Physical restraint training for nurses in English and Welsh Ó 2002 Blackwell Science Ltd, Journal of Advanced Nursing, 37(4), 394±401 Health and nursing policy issues psychiatric intensive care and regional secure units. Journal of Mental Health 10, 151±162. Lipsedge M. (1994) Dangerous stereotypes. Journal of Forensic Psychiatry 5, 14±19. Oakley P. (1998) Health Service Report. Summer 1998. Police Complaints Authority (2000) CS Spray: Increasing Public Safety? The Police Complaints Authority, London. Prosser D., Johnson S., Kuipers E., Szmukler G., Bebbington P. & Thornicroft G. (1997) Perceived sources of work stress and satisfaction among hospital and community mental health staff, and their relation to mental health, burnout and job satisfaction. Journal of Psychosomatic Research 43, 51±59. Reid Y., Johnson S., Morant N., Kuipers E., Szmukler G., Bebbington P., Thornicroft G. & Prosser D. (1999) Improving support for mental health staff: a qualitative study. Social Psychiatry and Psychiatric Epidemiology 34, 309±315. Rippon J.P. (2000) Aggression and violence in health care professions. Journal of Advanced Nursing 31, 452±460. Rosenthal T.L., Edwards N.B., Rosenthal R.H. & Ackerman B. (1992) Hospital violence: site, severity and nurses' preventive training. Issues in Mental Health Nursing 13, 349±356. Royal College of Psychiatrists Research Unit (1998) Management of Imminent Violence: Clinical Practice Guidelines to Support Mental Health Services (Occasional Paper OP41). Royal College of Psychiatrists, London. Southward R.D. (2001) Cutaneous burns from CS incapacitant spray. Medicine, Science, and the Law 41, 74±77. Steinmetz S.K. (1986) The Violent Family. In Violence in the Home Interdisciplinary Perspectives (Lystadad M. ed.), Brunner Mazel, New York, pp. 51±70. Thomas C., Bartlett A. & Mezy G.C. (1995) The extent and effects of violence among psychiatric in-patients. Psychiatric Bulletin 19, 600±604. Watts D. & Morgan G. (1994) Malignant alienation: dangers for patients who are hard to like. British Journal of Psychiatry 164, 11±15. Whittington R. & Wykes T. (1992) Staff strain and social support in a psychiatric hospital following assault by a patient. Journal of Advanced Nursing 17, 48±486. Wright S., Lee S., Sayer J., Parr A.M. & Gournay K. (2000) A review of the content of management of violence policies in inpatient mental health units. Mental Health Care 3, 373±376. Management of violence policy review Appendix Recommended items for inclusion in Trust management of violence policies De®nition of violence (specifying both physical and psychological harm) Statement of responsibility on the part of the authority Reference to MHA Code of Practice Statement of the aims of the policy Identi®cation of those responsible for ratifying, monitoring and evaluating the policy Date of policy and date of review Some account of the incidence of violence/threats Expectations and responsibilities of staff Commitment to appropriate training Mention of need for refresher training Intervals for refresher training speci®ed Preventative measures emphasized Potential causes of violence mentioned Information regarding warning signs of imminent violence Methods of coping: (a) de-escalation Methods of coping: (b) breakaways/self-preservation Methods of coping: (c) physical restraint Purpose of restraint Acceptable reasons for restraint given (See Royal College of Psychiatrists Research Unit 1998) Emphasis on physical restraint as measure of last resort Emphasis on use of minimum or reasonable force Mention of need to call for help Need for one staff member to take control of the incident mentioned Other patients to leave scene of the incident Mention of need for visual check for weapons Mention of unacceptable methods of restraint Need to maintain communication with the patient emphasized Reporting of incident in patient's case notes mentioned Reporting of incident for audit mentioned Circumstances under which police assistance should be sought described Advice on the use of CS spray by police in clinical areas Advice on the care and management of patients and staff exposed to CS spray Post-incident analysis and support for staff Post-incident analysis and support for other patients Post-incident analysis and support for assaultive patient Post incident review of care plan Occupational health/other staff welfare provision Information about support for legal help/compensation Outline of patient complaint procedure Gender issues People with physical and sensory handicaps (staff and patients) Communication and language differences/dif®culties Pregnancy (staff and patients) Seclusion protocol (see Royal College of Psychiatrists Research Unit 1998) Rapid tranquillization protocol (see Royal College of Psychiatrists Research Unit 1998) Nursing roles and responsibilities in relation to rapid tranquillization Ó 2002 Blackwell Science Ltd, Journal of Advanced Nursing, 37(4), 394±401 401