Search

Background Previous studies have found that patients with schizophrenia and bipolar disorder are more likely to be violent than the general population. The aim of this study was to investigate the association between schizophrenia and bipolar disorder and violent crime in the Israeli population. Methods Using the Israeli Psychiatric Hospitalization Case Registry we identified 3187 patients with a discharge diagnosis of schizophrenia and 506 patients with a discharge diagnosis of bipolar disorder. For each proband we identified parents and siblings, and gender- and age-matched controls for patients, parents and siblings. Information on violent crimes was obtained from police records. Results Patients with schizophrenia were at increased risk for violent crimes compared with controls [odds ratio (OR) 4.3, 95% confidence interval (CI) 3.8–4.9], especially women (OR 9.9, 95% CI 6.2–15.7). Risk for violent crimes was higher among patients with co-morbid substance misuse than in patients without such co-morbidity (OR 5.1, 95% CI 4.2–6.3). Patients with diagnosis of bipolar disorder were 2.5 times more likely to be convicted or released for mental reasons of violent crimes compared with controls and unaffected full siblings (OR=2.5, 95%CI 1.7–3.7, OR=2.5, 95%CI 1.6–4.0 respectively). Although men were more violent than women, diagnosis of bipolar disorder was a more significant risk factor for female patients than for male patients (OR=16.1 95%CI 1.8–144.6 vs. OR=2.4, 95%CI 1.5–3.7). Discussion The results of this study suggest that increased risk of violence is part of the clinical picture of schizophrenia and bipolar disorder and needs to be recognized as a legitimate, essential, aspect of clinical management.

The logistic regression proportional odds model is popular for analyzing studies with an ordered categorical outcome. In contingency table analysis, from a Type I error perspective, it is often thought best to collapse categories with sparse cell counts to improve asymptotic approximations used for testing hypotheses. Moreover, in the proportional odds model, it is natural to collapse adjacent categories of outcome since the slope parameter remains unchanged. This article asks the question: Is it really beneficial to do so? Using simulations, we show that in small samples collapsing categories produces Wald tests that are too conservative. Our simulations indicate that this is mainly due to stochastic dependence between the numerator and the denominator of the Wald statistic.

Although persons with schizophrenia (PWS) are entitled to health care and medical preventive and curative treatments in accordance with the same standards as other persons, they suffer from excess mortality compared with the general population. The main cause of natural premature death of PWS is attributed to cardiovascular disorders (CVD). We reviewed the studies of PWS, their risk factors and CVD mortality. In every study, PWS have increased risk of CVD mortality. Additionally, most but not all of the studies found increased CVD behavioral risk factors (i.e., smoking, sedentary life style/less physical activity, increased body mass index (BMI)) in PWS. In order to promote better health care to this population, we propose general recommendations to service users and their families, general and mental health professionals and policy makers. Severe mental illness such as schizophrenia should be addressed as a “risk equivalent” for CVD. Awareness of this association should be implemented in routine psychiatric and general medicine practice as well as at national levels for policy makers. Cardiovascular morbidity and mortality should be referred to as a measure of quality of care. Better communication between different physicians and other health care providers, who treat PWS, should be encouraged. This can be accompanied with technological advances (i.e., unified electronic medical record). Understanding of suggested treatment and adherence to recommendations in PWS may be improved by if relatives and friends accompany the patient in medical settings.