Schizophrenia and violence : a study of a one-year resident cohort of Scottish high security hospital patients with schizophrenia and their outcomes after ten years

• Main Author: Darjee, Rajan
• Published: University of Edinburgh 2011
• Subjects: 616.89
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.726455

INTRODUCTION: People with schizophrenia who commit serious violence are detained in secure hospitals. The research presented in this thesis investigated the long-term outcomes of patients with schizophrenia detained in a high security hospital. LITERATURE REVIEWS: Epidemiological research points to a significant but modest relationship between schizophrenia and violence. But when followed-up, offenders with schizophrenia are less violent than those with primary personality disorders or non-psychotic violent offenders. A number of factors have been found to be associated with violence in people with schizophrenia. Many of these are the same as factors associated with violence in non-psychotic populations. Seventy-two studies of the outcomes of UK security hospital studies published before 2010 were reviewed. Most included mixed diagnostic groups, and only 3 were focused on patients with schizophrenia. The most commonly reported outcome was conviction in over half of studies. Readmission, length of stay, institutional violence, the discharge process and mortality were reported in up to a quarter of studies. Very few studies reported clinical or social outcomes. AIMS AND RESEARCH QUESTIONS: The aims were to describe the long-term administrative (including moving though services and levels of security), forensic (including violence, conviction, absconding), clinical (including course of psychosis, treatment received, mortality) and social (including employment, intimate relationships, accommodation, substance misuse) outcomes of patients with schizophrenia detained in a high security hospital in Scotland; and to ascertain the factors associated with certain of these outcomes. Research questions were formulated to address these aims. METHOD: The sample was 169 patients with schizophrenia resident at the State Hospital, Carstairs, between August 1992 and August 1993. In 1992-4 patients and their psychiatrists were interviewed and data were collected from case records. Patients were followed up until the end of 2001. Records covering the intervening period were examined, and patients and independent informants were interviewed in 2000 and 2001. A number of structured instruments were used at interviews and to interrogate case records. Analysis involved descriptive statistics, survival analysis, bivariate approaches and logistic regression. RESULTS: Baseline characteristics: Most patients were male, unmarried and from socially disadvantaged backgrounds. The average patient was in their mid thirties. Most had been admitted from criminal courts after committing serious offences, half were subject to restriction orders and a quarter had killed. Most had had psychiatric treatment previously. Comorbid substance dependence and personality disorder were common. Administrative outcomes: Most patients (79%) left high security, but a minority (48% of those who left high security) reached the community, and 17% of patients who left high security were readmitted. Clinical correlates of outcomes were: not leaving high security - psychopathy and chronic psychosis; not reaching the community - chronic psychosis, not having substance dependence and not having an index offence precipitated by psychosis; readmission to high security - psychopathy. Other associates of outcomes, in terms of baseline and follow-up variables, were in keeping with the clinical correlates. Forensic outcomes: Three-quarters committed at least one act of violence, but less than a quarter caused serious harm. There were almost 2000 violent incidents, mostly in high security. Violence in the community was rare. The rate of criminal conviction was 13%, 10% for serious offences. Clinical correlates of forensic outcomes were: any violence - antisocial personality disorder and on going positive symptoms; serious violence - psychopathy; persistent violence - chronic positive and negative symptoms; convicted offending - psychopathy and substance dependence. Other associates were in keeping with these findings. Clinical and social outcomes: A third had episodic symptoms, a quarter continuous symptoms and a fifth recovery with no recurrence. Positive symptoms tended to improve whilst negative symptoms persisted. Persistent positive symptoms were related to not having psychopathy or substance dependence; and to more frequent relatively minor aggression, but not serious violence. The mortality rate was 7%. Social outcomes were very poor, but few patients were misusing substances at follow-up. Comorbid substance dependence and personality disorder: Substance dependent patients were more likely to be convicted, but otherwise had better administrative, clinical and social outcomes. Antisocial personality disorder was associated with violence during follow-up. Psychopathy was associated with not leaving high security, readmission to high security, conviction, serious violence and better outcome of psychosis. CONCLUSIONS: The outcomes and their associates are reviewed in light of the available literature. Course of psychosis and comorbid personality disorders and substance dependence are important clinical factors in determining the administrative, forensic and clinical outcomes of these patients. A typology based on these factors and course of violence is suggested as a way of synthesizing the findings with the literature and as a way of understanding the current findings. The methodological strengths and weaknesses of the study, and implications for clinical practice and research are discussed.