Pharmacological treatment and risk of psychiatric hospital admission in bipolar disorder.

Background Clinical trials have examined the efficacy of drugs to prevent relapse in patients with bipolar disorder, however, their design often limits generalisation to routine clinical practice. Aims To estimate the effectiveness of drugs used for maintenance treatment in bipolar disorder. Method We used national registers to identify 35 022 individuals diagnosed with bipolar disorder and information on lithium, valproate, carbamazepine, lamotrigine, quetiapine and olanzapine treatment from 2006 to 2009. The main outcome was psychiatric hospital admissions. We used stratified cox regression to compare periods on and off medication within the same individual. Results Medication with lithium, valproate, lamotrigine, olanzapine and quetiapine was associated with reduced rates of admission to hospital. Lithium was more effective than quetiapine and olanzapine. The effects of specific drugs depended on the polarity of the mood episode. Conclusions Our findings complement results from randomised controlled trails, but suggest that lithium is more effective than both quetiapine and olanzapine in routine clinical practice.