Completed suicides in 47 psychiatric hospitals in Germany--results from the AGATE-study.

INTRODUCTION There is an ongoing debate about a possibly enhanced risk of suicidal behaviour in some psychiatric patients due to psychopharmacotherapy. Our retrospective study aimed at analyzing the psychopharmacotherapy of 133 inpatient suicides and 133 controls by a matched pair design. METHODS We analyzed all suicides (n = 133) reported in the AGATE study from 1991 to 2008. Besides evaluation of sociodemographic variables and suicide methods, we compared psychopharmacotherapy of suicides with schizophrenia (n = 59) and affective disorders (n = 59) to that of a matched control group. RESULTS Most suicides (n = 102, 76.7%) were judged not to be related to psychopharmacotherapy. In general, more psychopharmacological drugs were prescribed for suicides than for controls. Schizophrenic suicides received more low potency FGAs than their controls. More suicides with affective disorders than controls were treated with NASSAs, SNRIs, and high or low potency FGAs. In contrast to their controls, none of the suicides with affective disorders received lithium. NASSAs, SNRIs, and high or low potency FGAs predicted suicide in regression analysis for inpatients with affective disorders. DISCUSSION Differences in psychopharmacotherapy might mainly result from a recognized risk of suicide or a more severe degree of illness. However, the underrepresentation of lithium in the suicide groups is noticeable.