Reducing violence in severe mental illness

BMJ 2001; 323 doi: https://doi.org/10.1136/bmj.323.7321.1080 (Published 10 November 2001) Cite this as: BMJ 2001;323:1080

Community care does not do well

  1. Tilman Steinert, head of research department and clinical department for general psychiatry (Tilman.Steinert{at}zpf-weissenau.de)
  1. Centre of Psychiatry, University of Ulm, D88214 Weissenau, Germany

    Papers p 1093

    Two years ago, Munk-Jørgensen initiated a continuing debate about the development of psychiatric care for severely mentally ill people in a paper entitled “Has deinstitutionalization gone too far?”1 He pointed out that the reduction in numbers of psychiatric hospital beds had been accompanied by a continuing increase in the number of forensic psychiatric patients as well as an increase in suicides and readmissions in Denmark. Similarly, Webster et al reported a doubling of the number of forensic patients within the past decade in Canada.2 In the USA, meanwhile, a considerable proportion of severely mentally ill people do not live within the community but are imprisoned3 or homeless.4 The study by Walsh et al in this issue (p 1093) is the first randomised controlled study to examine whether an increased intensity of psychiatric community care can reduce violence among severely mentally ill patients managed in the community.5 Its results are disappointing.

    A considerable proportion of patients in this study—more than one in five—carried out physical attacks both in the intensive intervention group and in the standard care group. The intensive care group had received over twice as many contacts (which included those related to medication and to the criminal justice system). Before we draw harsh conclusions about the ineffectiveness of outpatient psychiatric care, however, we should keep in mind one limitation of Walsh et al's study. They recorded whether patients ever committed physical attacks during the observation period, but frequency and severity of violence were not recorded. Thus, it is possible that the frequency and severity of the violence were reduced in the experimental group, even though the number of patients committing violent acts was not. Nevertheless, we are still confronted with the politically meaningful question of whether our present forms of outpatient care are adequate for certain groups of psychiatric patients.

    In their discussion Walsh et al raise the question whether assertive community treatment and intensive case management need modifying to be more effective in patients prone to violent behaviour. To date there is not much evidence on the form these modifications would have to take. Actually, a rather effective treatment for violence in seriously mentally ill people is available, but only for limited periods: hospital treatment. Our group has shown a continuous day by day reduction in the incidence of violent behaviour among inpatients with schizophrenia, resulting in very low rates after some weeks of treatment.6

    Violence in acute psychiatric units is strongly associated with the severity of psychotic symptoms, while factors such as substance abuse, non-compliance with medication, criminal peers, and poor living conditions are minimised under the regimen of psychiatric wards, which can be locked and where medication and drug abstinence can be enforced. Conditions in the community are quite different. In a systematic meta-analysis of research on predictors of criminal and violent recidivism among mentally disordered offenders Bonta et al found that predictors of violence in people with major mental disorders are nearly the same as those in people without such disorders: criminal history, age, substance abuse, deviant lifestyle, family problems, antisocial personality disorder.7 These findings are confirmed by the results of Walsh et al. In contrast to these so called “static” variables, “dynamic” variables such as psychopathology and clinical assessments were identified as only weak predictors of violent behaviour in the community.8

    From this point of view, therefore, it is no surprise that measures of therapeutic care in the community do not yield substantial results in what is more a problem of general crime prevention than of mental illness. As Walsh et al emphasise, further research should address the question of whether forms of compulsory outpatient treatment combined with psychosocial support can be developed. These need to be effective in reducing violence in a core group of mentally disordered people.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    View Abstract

    Sign in

    Log in through your institution