Management of Risk in Psychiatric Rehabilitation

John McGovern looks at factors to consider when assessing patients' risk of self harm and harm to others

The Psychologist, September 1996, ps 405-408

For psychologists involved in making decisions, on which patients are suitable for admission to/discharge from hospital psychiatric rehabilitation units, questions of self harm or harm to others are vital issues. In these units more emphasis is placed on enabling patients to achieve the optimum level of independence, in contrast to the close monitoring and security of the acute and secure units from which many of them are referred. Making decisions on levels of risk and strategies of management is not a new phenomenon (e.g. Thompson, 1995), but with the present government policy of returning individuals to live in the community whenever possible (HMSO, 1975, 1989a 1989b), the need for such decisions is ever more frequent. In this article I will examine the ability of clinicians to make accurate judgements about patients risks of self-harm or harm to others. will review the research evidence on predictive factors, to see whether the findings can assist clinicians manage risk more effectively.

In-patient changes

With regard to the management of risk of self-harm or harm to others in psychiatric rehabilitation settings, the main concern when making decisions on admission or discharge are the risks of suicide or violence to others. There are a number of reasons why these two factors are of increasing concern to clinicians in psychiatric rehabilitation units. These are:

1) Changes in the characteristics of patients treated in these units, i.e. in contrast to the 'old long-stay patient' whose problems were predominantly life skill deficits due to illness factors and institutionalisation, current inpatients generally have relatively high levels of functioning, but have poorly controlled symptomatology and show disruptive behaviour. This latter group, who remain in hospital despite attempts to discharge them, suffer frequent relapses and require re-admission and have bee described as the 'new long-term (Shepherd, 1991; Lelliot, Wing Clifford, 1994).

2) Research findings on the vulnerability of patients to suicide in the first four months following discharge (Flood & Seager, 1968; Roy, 1982 Crammer, 1984; Appleby, 1992).

3) Government proposals and change in discharge protocol (Department of Health, 1993) in the light of high exposure, media coverage (Barnes & Earnshaw, 1993) and mounting public concern (Link, Andrews & Cullen, 1992; Weddle, 1995) over incidents of suicide and murder by ex-hospital psychiatric patients. The report of the Christopher Clunis Inquiry (Richie, Dick & Lingham, 1994) focused on mental health services and their failure to provide an adequate safety net.

In order to understand the severity of mental health problems of patients in these units, let us examine a snapshot of the in-patients of the Mental Health Rehabilitation Service in two community units in Forth Valley, Central Scotland from November 1995. Figure I shows a table detailing the characteristics of patients treated in these units at present. In line with these aspects many of these patients could be considered 'the new long-term' and as such carry some of the risks of self-harm and harm to others associated with this group.

In a national audit of new long-term psychiatric patients in Scotland, Ireland, England and Wales (Lelliot et al., 1994) the researchers found that half of the patients were considered by the assessors to pose a moderate or severe risk of violent or self destructive behaviour were they to be discharged. Researchers have found conflicting evidence on the accuracy of clinicians' ability to predict violence by patients (Monahan, 1981; Werner, Rose, Yesavage & Seeman, 1984; Lidz, Mulvey & Gardner, 1993). It is crucial, therefore, that clinicians become aware of what specific factors have been found to be predictors of these risks and how reliable or valid these predictions are, With regard to the assessment of the risk of violence by psychiatric patients, there are numerous factors which have been investigated. The key areas of clinical promise are anger control, psychopathy, impulsiveness and psychotic symptoms.

Figure 1: Characteristics of psychiatric rehabilitation in-patients in Forth Valley, Central Scotland, November 1995 (N = 21)

Schizophrenic Affective Disorder Other

Depressive-Manic

N = 12 N = 8 N = I

Mean Range Mean Range Mean

Length of illness (years) 16.6 1 - 43 12.8 2 - 26 6

Time as in-patient (years) 4.1 0.06-19.5 1.2 0.1-5.3 0.25

Length of present admission (months) 24.3 3-94 12 3-28 3

Parasuicide attempts 2 0-15 1.5 0-5 18

Number %of total Number % of total

Caused injury to others 3 25 4 50 -

Active hallucination 7 58.3 2 25 -

Active delusions 10 83.3 4 50 -

Alcohol/drug abuse problems 2 16.6 3 37 1

Number currently detained on Section 5 41.6 1 12.5 -

Number spent time in Special Hospital 2 16.6 0 0 -

Predictors of risk of violence

Anger

Anger was found by Kay, Wolkenfeld and Murrill (1988) to be the strongest predictor of physical aggression in the clinical and diagnostic profiles of their cohort of 208 psychiatric patients. Other researchers have found associations between anger (Craig, 1982), irritability (Segal, Watson, Goldfinger & Averbuck, 1988), dangerousness to others, and assaultive behaviour with psychiatric admission patients.

Using rating scales designed to measure clinical levels of anger, Novaco (1994) found an association with physical assault, i.e. 3.8 per cent of 'low anger' patients were assaultive in contrast to 28.5 per cent of the 'high anger' patients. A prospective analysis which examined 1987 anger with 1988 assaultiveness found that 'high anger' patients were still three times more likely to be assaultive than 'low anger' patients. Based on his seminal work on understanding anger (Novaco, 1976, 1977), Novaco highlights the inability of present scales, which view anger as a normal emotion, to provide a differentiated assessment of the dysfunctional aspects of anger and the relationship of cognitive, behavioural and arousal level components of it, with aggression. To address these deficiencies he has developed the Novaco Anger Scale (parts A and B). The second component (part B) is an abbreviated improvement of the Novaco Provocation Inventory (1988), providing an index of anger intensity and generality across a range of potentially provocative situations. Retrospective analysis with the new scale has shown correlations with a wide range of aggressive behaviour criteria and, perhaps most importantly, with the number of convictions for violent crimes against others (r = .34 with NAS part A). Despite the complexity of risk management and its relationship with anger, the concurrent and predictive validity of the Novaco Anger Scale in comparison with other indices of aggression (Novaco, 1994) suggest a constructive role for it in clinical practice and future research. A further measure, which may have some specific utility in screening decisions in outpatient intake procedures, is the six item Brief Anger-Aggression Questionnaire (Maiuro, Vitaliano & Cahn, 1987).

Impulsiveness

The potential importance of impulsiveness, which is part of the behavioural domain of the Novaco Anger Scale (Novaco, 1994), as a predictor of risk, has been investigated by Barratt (1972). Results from a study using the Barratt Impulsiveness Scale - 10 (Barratt, 1994) with adults, students, offenders and psychiatric in-patients found that the latter scored highest on non-planning impulsiveness (i.e. a lack of concern for the future). Among clinical populations, patients with substance abuse problems, antisocial personality disorders and impulsive aggressive tendencies tend to score high on impulsiveness overall. Barratt (1994) hypothesises that impulsiveness is part of an action-orientated second order personality trait significantly related to one form of aggression, i.e. impulsive aggression. This form of aggression is characterised by someone with a 'hair trigger temper' who acts without thinking and, who during the act does not process incoming stimuli logically. High impulsive subjects in general are reported as being 'present oriented', more unreliable in making appointments and not having stable long-lasting interpersonal relationships. In support of his hypothesis, Barratt cites behavioural laboratory studies, clinical observations, biological research on serotonin levels (Brown, Kent, Bryant, Gevedon, Campbell, Felthous, Barratt & Rose, 1989; Kent, Brown, Bryant, Barratt, Felthous & Rose, 1988) and genetic inheritance (Pedersen, Plomin, McLearn & Friberg, 1988). The value of impulsiveness as a valuable predictor of risk with psychiatric patients will depend on future specific research findings on associations with impulsiveness and acts of self harm or aggressive behaviour in prospective studies.

Psychopathy

Psychopathy, which has impulsivity as one of its features, has been the subject of another important area of risk assessment research. A number of studies have looked at the associations between psychopathy, as assessed using the Psychopathy Checklist (PCL) and PCL - Revised (R) (Hare, Hart & Harpur, 1991), schizophrenia and violence. The PCL - R consists of 20 items designed to assess a range of relevant personality traits and behaviours. In a long-term (10 years on average) follow-up study of 169 male patients released from a maximum security psychiatric hospital, Harris, Rice and Cormier (1991) found that PCL - R scores were strongly correlated r = .42 with violent recidivism. Heilbrun, Hart, Hare, Gustafson, Nunez and White (1993) looked at the association between psychopathy, schizophrenia and violence in a sample of 218 male consecutive admissions to a state forensic hospital in Florida. In-patient violence was indexed by verbal and physical assault during the first and the last three months of hospitalisation. For the 183 patients who were subsequently released from hospital, Heilbrun et al. also coded violent recidivism, that is, any arrest for a violent offence that appeared on the patient's Federal Bureau of Investigation criminal record during the follow-up period. The PCL - R was correlated r = .30 with assaults during the first three months of hospitalisation, but only r<0.03 with assaults in the last three months. The PCL - R was also correlated with violent recidivism. Hart, Hare and Forth (1994) report on a revised version of this very promising risk assessment measure, the PCL - Screening Version, which is shorter to administer and is appropriate for use in non-forensic settings. Although much of the promising risk assessment research has focused on personality variables, investigators have also begun to look into the question of whether various symptoms of mental disorder are associated with an increased risk of violence towards others.

Psychotic symptoms

A National Institute of Mental Health epidemiological survey conducted over five communities in the USA found that people who assaulted others were significantly more likely to have a psychiatric disorder. In general, mental illness alone was about twice as prevalent in the sub group identified as violent, while those with substance abuse alone and co-morbidity were about five times more prevalent among those who reported violence than among those that did not. Among people who had never been arrested and were not currently mentally ill, merely being a former mental patient was not significant risk factor for violent behaviour in the community.

A recent study (Link et al., 1992) illustrated eloquently a moderating relationship between the occurrence of psychotic symptomatology and violence. When they controlled for current psychotic symptomatology they found that differences on measures of violence between patients and a community sample disappeared. Some preliminary work has been done on psychotic delusions as a predictor of risk. In order to study different aspects of delusional experience the Maudsley Assessment of Delusions Schedule was created. Reporting on its use in predicting violence, Taylor, Garety, Buchanan, Reed, Wessely, Katarzyna, Dunn and Grubin (1994) found that 11 per cent (9) of the 83 general psychiatry patients so far studied using this schedule had acted violently even towards themselves, but only one 28-day period was studied.

The other specific psychotic symptom which has attracted attention as a risk marker is auditory hallucinations, particularly command hallucinations. Studies have shown that although patients may respond to command hallucinations, no significant association has been found with assaultive behaviour (Hellerstein, Frosch and Koenigsberg, 1987; Rogers, Gillis, Turner & Frise-Smith, 1990).

Overall, studies based on quantitative clinical ratings of psychotic patients have shown a significant positive relationship between hallucinations and violence together with correlations with other positive psychotic symptoms such as thinking disturbance and conceptual disorganisation. It would appear, therefore, that psychotic symptoms would join anger, impulsiveness and psychopathy on a list of key predictive risk factors of violence to others which require further research.

Risk of self-harm

Patient characteristics

Research on risk of self harm to others also appears to be at an early stage of development. Much of the work carried out has involved retrospective analysis of suicides identified by the coroner where those involved have a previous psychiatric history and where case notes are examined/rated and primary clinicians interviewed. In terms of managing the risk for self harm a number of important findings emerged. A history of parasuicide was a risk factor in 50 percent of cases (Myers & Neal, 1978; R 1982) and Rorsmann (1973) reported th factor as increasing ten-fold the risk suicide. The majority of the retrospective studies highlighted dangerous periods the beginning and end of an episode illness, particularly the latter, with substantially increased risk in the three months following discharge. In the majority of cases those who committed suicide had been seen by a clinician no more than a month earlier, and in a substantial number clinical improvement had been observed despite unresolved social or interpersonal problems.

Statistics on suicide risk for diagnostic categories vary in the research reports. Allenbeck and Wistedt (198 established that 3.9 per cent of 1,190 schizophrenics discharged in a single year and followed up for 10 years were officially recorded as having died by suicide or an undetermined cause. Fawcett Scheftner, Clark, Hedeker, Gibbons and Coryell (1987) came up with a similar figure of 3.6 in a 10-year follow-up of major affective disorder including schizoaffective disorder.

Treatment factors

In addition to patient characteristics, number of studies have looked at treatment aspects. Morgan and Priest (1984 1991) suggest that suicidal risk patients should not be discharged until situational factors have been addressed and that mere improvement in reported symptoms is an insufficient discharge criterion. These authors point to the fact that in half the cases of suicide by in patients, significant improvements in clinical state had been observed. The suggest that this reported improvement signals a resolution of internal conflict in favour of self-harm rather than reduced suicide risk. Other risk factors particularly related to increased opportunity for suicide which have been cited are low staff numbers, little structure in ward programme, alienation from staff members and poor design in treatment settings (Crammer, 1984; Morgan & Priest, 1991; Appleby, 1992).

Vulnerability factors

In contrast to retrospective studies, Cohen, Test and Brown (1990) carried out a prospective study of eight schizophrenic young men who committed suicide during an ongoing longitudinal study. Baseline measures of self-reported subjective distress were consistently predictive of later suicide, whereas interviewer rated measures and post-baseline assessments of distress were not. This finding is consistent with Beck, Steer, Kovacs and Carrison's (1985) report, that patients who eventually committed suicide had significantly higher hopelessness scores than those who did not, when they were assessed during hospitalisation several years before their deaths. This suggests an underlying schema or cognitive set which lies dormant until activated by some stressful event and then becomes a vulnerability factor for suicidal risk. The important role of hopelessness in schizophrenia is given further credence by Drake and Cotton's (1986) finding that pre-suicidal schizophrenics experience depressed mood, but that the relationship between depression and suicide disappears when hopelessness was taken into account.

This theme of hopelessness is reflected by Appleby's (1992) review of suicide amongst psychiatric patients in that, in schizophrenia those most at risk are likely to be young, have a short illness, have a history of parasuicide, and features of affective disorders particularly hopelessness. They are also likely to be unemployed, unmarried and have expressed suicidal ideas during their last admission. These factors together with an awareness of treatment aspects, ward operation and periods of high risk following discharge are all pertinent to developing a risk management strategy.

Conclusion

Management of risk, of self harm or harm to others in a psychiatric rehabilitation unit providing a service predominantly to the 'new long-term' is dependent on an awareness of what factors might increase or decrease risk. From the present analysis a number of predictive elements of risk have emerged from the research literature. The confidence though with which these findings can be interpreted is limited by a number of reservations. Much of the research to date is gained from retrospective analysis which has problems in terms of clinicians memory and possible bias after traumatic events. Measures of the phenomenon of violence, which is basically an antisocial and illegal behaviour, may also be subject to a range of possible biases when dependent on arrest/hospital admission data or self-report. Furthermore, given the complexities of the areas being studied, there is much reliance on correlation data rather than causal relationships. The extent to which the findings of this research can be generalised to the subgroups of psychiatric rehabilitation patients is also open to question, as is the relative weighting clinicians can give to specific predictors of risk or indeed interactions of these predictors.

Improving risk management

It may be that in line with the present research on patients' individual relapse signatures (Birchwood & Tarrier, 1994) clinicians should look at individual risk vulnerability profiles. The first step towards doing this would be to use pre- post- and follow-up measures of anger, impulsivity, psychopathy, individual psychotic symptoms, and hopelessness Where hopelessness is detected, appropriate risk management may be to offer the patient cognitive therapy as proposed by Beck et al. (1979) to modify underlying vulnerability to risk of self harm. Introducing these measures to a well staffed and structured in-patient Psychiatric Rehabilitation Unit, which focused on assisting residents to gain confidence in managing their symptoms and resolve social problems prior to discharge, will be heading in the direction of effective risk management, while building a valuable data base for prospective studies. These measure would complement our existing clinical assessments of risk, based on observed/recorded incidents of violence, self-harm, suicidal ideation and emergency medication (particularly in the last month/s) and direct observation of level of disturbance of a person' mental state during a selection/progress review interview. For the 'scientist practitioner' clinical psychologist who may be involved in carrying out these interviews, the adoption of these formal measures would provide a degree of comfort in an area where clinical experience has demonstrated the fallibility of clinical information. Appropriate intermediate steps to full discharge in close liaison with those providing out-patient continuing care may offer individuals a degree of support, which will help them cope during this vulnerable transition period and diminish considerably risk of self harm or harm to others.

Acknowledgements

The author wishes to thank Elaine Carr and Jenny Marsden for their work in collecting and presenting the data on the characteristics of Mental Health Rehabilitation in-patients in Forth Valley and to the staff in the Rehabilitation Units who assisted them in this task. Gratitude is also expressed to Audrey Caw McGovern, Lorraine Phillip and Dr Simon Thompson for their proof-reading of the paper and their helpful comments.

References

Allenbeck, P. & Wistedt, B. (1986). Mortality in schizophrenia. A ten-year follow-up based on the Stockholm County in-patients register- Archives General Psychiatry, 43, 650-653.

Appleby, L. (1992). Suicide in psychiatric patients: risk and prevention. British Journal of Psychiatry, 161, 749-758.

Barnes, R.C. & Earnshaw, S. (1993). Mental illness in British newspapers. Psychiatric Bulletin, 17, 673-674.

Barratt, E.S. (1972). Anxiety and impulsiveness: toward a neuropsychological model. In C. Spielberger (Ed.): Anxiety: Current Trends in Theory and Research. New York: Academic Press.

Barratt, E.S. (1994). In J. Monahan & H.J. Steadman (Eds): Violence and Mental Disorder: Developments in Risk Assessment. Chicago: The University of Chicago Press.

Beck, A.T., Rush, J., Shaw, B. & Emery, C. (1979). Cognitive Therapy of Depression. New York: Guilford Press.

Beck, A.T., Steer, R.A., Kovacs, M. & Garrison, B. (1985). Hopelessness and eventual suicide: a 10-year prospective study of patients hospitalised with suicidal ideation. American Journal of Psychiatry, 142, 559-563.

Birchwood, M. & Tarrier, N. (Eds) (1994). Psychological Management of Schizophrenia. Chichester: John Wiley & Sons Ltd.

Brown, C., Kent, T., Bryant, S., Gevedon, R., Campbell, J., Felthous, A., Barratt, E. & Rose, R. (1989). Blood platelet uptake of serotonin in episodic aggression. Psychiatry Research, 27, 5-12.

Cohen, L.J., Test, M.A., & Brown, R.L. (1990). Suicide and schizophrenia: data from a prospective community treatment study. American Journal Psychiatry, 147, 602-607.

Craig, T.J. (1982). An epidemiological study of problems associated with violence among psychiatric in-patients. American Journal of Psychiatry, 139, 1262-1266.

Crammer, J.L. (1984). The special characteristics of suicide in hospital in-patients. British Journal of Psychiatry, 145, 460-476.

Department of Health (1993). Legal Powers on the Care of Mentally Ill people in the Community. Report the Internal Review. London: Department of Health.

Drake, R.E. & Cotton, P.G. (1986). Depression, hopelessness and suicide in chronic schizophrenia. British Journal of Psychiatry, 148, 554-559.

Fawcett, j., Scheftner, W., Clark, D., Hedeker, D., Gibbons, R. & Coryell, W. (1987). Clinical predictors of suicide in patients with major affective disorders: a controlled prospective study. American Journal of Psychiatry, 144, 35-40.

Flood, R.A. & Seager, C.P. (1968). A retrospective examination of psychiatric case records of patients who subsequently committed suicide. British Journal of Psychiatry, 114, 443-450.

Hare, R.D., Hart, S.D. & Harpur, T.J. (1991). Psychopathy and the proposed DSM IV criteria for antisocial personality disorder. Journal of Abnormal Psychology, 100, 391-39.

Harris, G.T., Rice, M.E., Cormier, C.A. (1991). Psychopathy and violent recidivism. Law and Human Behaviour, 15, 625-637.

Hart, S.D., Hare, R.D. & Forth, A.E. (1994). Cited in J. Monahan & H.J. Steadman (Eds): Violence and Mental Disorder: Developments in Risk Assessment. Chicago: The University of Chicago Press.

Heilbrun, K., Hart, S.D., Hare, R.D., Gustafson, D., Nunez, C. & White, A.J. (1993). Inpatient and post-discharge aggression in mentally disordered offenders: The role of psychopathy. Unpublished manuscript. Cited in Chapter 4, J. Monahan & H,J, Steadman (Eds): Violence and Mental Disorder: Developments in Risk Assessment. Chicago: The University of Chicago Press.

Hellerstein, D., Frosch, W. & Koenigsberg, H.W. (1987). The clinical significance of command hallucinations. American Journal of Psychiatry, 144, 219-221.

HMSO (1975). Better Services For the Mentally Ill. London: HMSO.

HMSO (1989a). Caring for People. London: HMSO.

HMSO (1989b). Working For Patients. London: HMSO.

Kay, S.R., Wolkenfeld, E & Murrill, L.M. (1988). Profiles of aggression among psychiatric inpatients II: Covariates and Predictors. The Journal of Nervous and Mental Disease, 176, 547- 557.

Kent, T., Brown, C., Bryant, S., Barratt, E., Felthous, A. & Rose, R. (1988). Blood platelet uptake of serotonin in episodic aggression: correlation with red blood cell proton TI and impulsivity. Psychopharmacology Bulletin, 24, 454-457.

Lelliot, P., Wing, J. & Clifford, P. (1994). A national audit of new long stay psychiatric patients. 1: method and description of cohort. British Journal of Psychiatry, 165, 160-169.

Lidz, C., Mulvey, E. & Gardner, W. (1993). The accuracy of predictions of violence to others. Journal of the American Medical Association, 269, 1007-101 1.

Link, B.G., Andrews, H.A. & Cullen, F.T. (1992). The violent and illegal behaviour of mental patients reconsidered. American Sociological Review, 57, 275292.

Maiuro, R.D., Vitaliano, PP. & Cahn, T.S. (1987). A brief measure for the assessment of anger and aggression. Journal of Interpersonal Violence, 2, 166-178.

Monahan, J. (1981). The Clinical Prediction of Violent Behaviour. Washington, DC: Government Printing Office.

Morgan, H.C. & Priest, P. (1984). Assessment of suicide risk in psychiatric in-patients. British Journal of Psychiatry, 145, 467-469.

Morgan, H.C. & Priest, P. (1991). Suicide and other unexpected deaths among psychiatric in-patients. The Bristol Confidential Inquiry. British Journal of Psychiatry, 158, 368-374.

Myers, D.H. & Neal, C.D. (1978). Suicide in psychiatric in-patients. British Journal of Psychiatry, 133, 38-44.

Novaco, R.W. (1976). The function and regulation of the arousal of anger. American Journal of Psychiatry, 133, 1124-1128.

Novaco, R.W. (1977) Stress inoculation: a cognitive therapy for anger and its application to a case of depression. Journal of Consulting and Clinical Psychology, 45, 600-608.

Novaco, R.W. (1988). Novaco Provocation Inventory. In M. Hersen & A. Bellack (Eds): Directory of Behaviour Assessment Techniques. New York: Pergamon Press.

Novaco, R.W. (1994). Anger as a risk factor for violence among the mentally disordered. In J. Monahan &- H.J. Steadman (Eds): Violence and Mental Disorder: Developments in Risk Assessment, Chicago: The University of Chicago Press.

Pedersen, N., Plomin, R., McLearn, G. & Friberg, L. (1988). Neuroticism, extroversion, and related traits in adult twins reared apart and reared together. Journal of Personality and Social Psychology, 55, 950-957.

Richie, J., Dick, D. & Lingham, R. (1994). Report of the Christopher Clunis Inquiry North West Thames Regional Health Authority.

Rogers, R., Gillis, J. R., Turner, R.E. & Frise-Smith, T. (1990). The clinical presentation of command hallucinations in a forensic population. , American Journal of Psychiatry, 147, 1304 - 1307.

Rorsmann, B. (191-3). Suicide in psychiatric patients: A comparative study. Social Psychiatry, 8, 55-66.

Roy, A. (1982). Suicide in chronic schizophrenia. British Journal of Psychiatry, 141, 613-617.

Segal, S.P., Watson, A., Goldfinger, S.M. & Averbuck, D.S. (1988). Civil commitment in the psychiatric emergency room II: Mental disorder indicators and three dangerousness criteria. Archives of General Psychiatry, 45, 753-758.

Shepherd, G. (1991) Foreword: Psychiatric rehabilitation for the 1990s. In F.N. Watts; &- D.H. Bennett (Eds): Theory and Practice of Psychiatric Rehabilitation. Chichester: John Wiley and Sons Ltd.

Taylor, P.J., Garety, P., Buchanan, A., Reed, A., Wessely, S., Katarzyna, R., Dunn, G. & Grubin, D. (1994). Chapter 7, cited in J. Monahan &- H.J. Steadman (Eds): Violence and Mental Disorder: Developments in Risk Assessment. Chicago: The University of Chicago Press.

Thompson, S.B.N. (1995). Night nurses' paralysis. The Lancet, 346, 8978: 848.

Weddle, I. (1995). No loonies here. The Big Issue in Scotland, 15 Nov, 22-23.

Werner, P., Rose, T., Yesavage, J. & Seeman, K. (1984). Psychiatrists' judgements of dangerousness in patients on an acute care unit. American Journal of Psychiatry, 141, 263-266.

John McGovern is Chair/Co-ordinator of the Mental Health Rehabilitation Services, Central Scotland Health Care Trust, Bellsdyke Hospital, Larbert, Scotland FK5 4SF