Improving the design of maintenance studies for bipolar disorder

Background: In contrast to the trial design of acute mania studies, there is no standard design for bipolar maintenance studies. Over the past 15 years, the design of monotherapy maintenance studies in bipolar disorder has evolved significantly, but recent study designs continue to differ in important ways. Scope: We reviewed the design of recent controlled bipolar maintenance studies, using PubMed, from August 2006 to August 2009, examining the strengths and weaknesses of different study design features. Findings: Design differences are sufficiently important that the disparate results across maintenance studies may reflect either true differences in medication efficacy or the effects of these design differences on outcome. Design elements such as recent episode polarity, stabilization criteria, using enriched versus nonenriched samples, length of stabilization before randomization, length of experimental phase, and recurrence outcome criteria are critical factors that differ widely across studies and likely play a role in study outcome. Conclusions: As consensus for trial designs for bipolar maintenance therapy is developed, it will be easier to develop algorithms for maintenance treatment based on results from studies as opposed to clinical opinions.