Determinants of victimization in patients with severe mental illness: results from a nation-wide cross-sectional survey in the Netherlands.

OBJECTIVE

To examine determinants of the prevalence and frequency of criminal victimization (i.e. both personal and property crime victimization) in outpatients with severe mental illness.

METHODS

Data was collected using a multisite epidemiological survey including a random sample of 956 adult outpatients with SMI. Data were collected between 2010 and 2012. Data on 12-month victimization prevalence and frequency were obtained using the victimization scale of the Dutch Crime and Victimization Survey. Demographic characteristics, clinical determinants, e.g., clinical diagnosis, psychosocial functioning, drug use and alcohol abuse over the past 12 months, co-morbid PTSD diagnosis, and victimological determinants e.g. physical abuse, physical neglect and sexual abuse in childhood, perpetration of violence over the past 12-months, and anger disposition were assessed as determinants. Univariable and multivariable hurdle regression analyses were conducted to test associations of the potential determinants with victimization prevalence and frequency.

RESULTS

Twelve-month prevalence rates of personal and property crime were 19% and 28%, respectively. Clinical characteristics were more pronounced regarding personal crime victimization. In the multivariable model, presence of psychotic disorder, drug use, childhood physical and sexual abuse, and recent violent perpetration were associated with the 12-month prevalence or frequency rate of personal crime victimization. Native Dutch and divorced patients were more at risk as well. Next to this being employed, poor social functioning, having perpetrated a violent crime, as well as alcohol abuse and recent drug use were all significantly related to property crime prevalence or frequency rate in the multivariable model. In absolute terms, the effect sizes observed tended to be moderate to small. The multivariate models, however, explained the outcome variance moderately well (Nagelkerke's pseudo R = 25.0 - 27.9%.

CONCLUSIONS

Clinicians should be aware of the high risk of victimization among their patients with severe mental illness. Particular attention should be devoted to people with substance use histories and perpetrators of violence, since they are also at an increased risk of being victims as well.