Background: The aim of this study was to estimate the proportion of homicide recidivists among population studies of homicide offenders with schizophrenia.Methods: Systematic review and meta-analysis of published studies of homicide associated with schizophrenia conducted in defined populations and indexed in Medline, PsychINFO, or Embase between January 1960 and November 2013. Published data was supplemented with unpublished data about recidivism obtained by personal communication from the authors of published studies of homicide and schizophrenia. Random effects meta-analysis was used to calculate a pooled estimate of the proportion of homicide recidivists.Results: Three studies reported that 4.3%, 4.5%, and 10.7% of homicide offenders with schizophrenia had committed an earlier homicide. Unpublished data were obtained from the authors of 11 studies of homicide in schizophrenia published in English between 1980 and 2013. The authors of 2 studies reported a single case of homicide recidivism and the authors of 9 studies reported no cases. The rates of homicide recidivism between studies were highly heterogeneous (I-square = 79). The pooled estimate of the proportion of homicide offenders with schizophrenia who had committed an earlier homicide was 2.3% (95% CI (Confidence Interval) 0.07% to 7.2%), a figure that was not reported in any individual study. The pooled proportion of homicide recidivists from published reports was more than ten times greater (8.6%, 95% CI 5.7%-12.9%) than the pooled proportion of homicide recidivists estimated from data provided by personal communication (0.06%, 95% CI 0.02% to 1.8%).Conclusions: In most jurisdictions, homicide recidivism by people with schizophrenia is less common than published reports have suggested. The reasons for the variation in the rates of homicide recidivism between studies are unclear, although in most jurisdictions long-term secure treatment and supervision after release appears to be effective in preventing homicide recidivism. A prospective study conducted in a large population or in multiple jurisdictions over a long period of time might result in a more accurate estimate the risk of a second homicide by a person with schizophrenia.
Bibliographical noteCopyright the Author(s) 2014. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.
- Repeat homicide