To Örebro University

Background: Electroconvulsive therapy (ECT) is effective in bipolar depression, but relapse is common. The aim of the study was (i) to identify prognostic factors (ii) and to determine the impact of pharmacological approaches on the risk for rehospitalization or suicide.

Methods: This register study analyzed data from individuals treated with inpatient ECT for bipolar depression. Subjects were identified using the Swedish National Patient Register between 2011 and 2014 and the Swedish National Quality Register for ECT. Other national registers provided data on psychopharmacotherapy, socio-demographic factors, and causes of death. The endpoint was the composite of rehospitalization for any psychiatric disorder, suicide attempt or completed suicide (RoS). Cox regression was used to calculate hazard ratios in univariate and multivariate models.

Results: Data from 1255 patients were analyzed. The mean period of follow-up was 346 days.

A total of 29%, 41%, and 52% of patients reached RoS at 3, 6, and 12 months post-discharge. A history of multiple psychiatric admissions, lower age, and post-discharge treatment with antipsychotics or benzodiazepines was associated with RoS.

Limitations: Indication bias may have affected the results.

Conclusions: A history of multiple hospital admissions and lower age are key predictors of the composite of rehospitalization or suicide in patients treated with ECT for bipolar depression. Lithium might be effective. By contrast, antipsychotics and benzodiazepines were associated with increased risk, but possibly this finding was influenced by indication bias.