Optimizing treatment outcomes in bipolar disorder under ordinary conditions.

The gold standard for establishing the efficacy of a drug is the randomized controlled trial (RCT). In the past decade, there has been an increasing recognition of the important difference between RCTs designed to demonstrate efficacy and treatment studies designed to evaluate the effectiveness of various therapies in actual clinical practice. This distinction is especially true for bipolar disorder, a complex illness whose treatment response is significantly influenced by various moderating baseline factors (e.g., comorbidity, severity, treatment resistance) and mediating factors (e.g., adequacy of dose and treatment duration, use of concomitant medications, treatment adherence). To reduce sources of variance in treatment outcome, RCTs rely on restrictive entry criteria in an attempt to control moderating factors and on highly standardized study procedures in an attempt to control mediating factors. The goal is to reduce the heterogeneity in the patient sample and in treatments administered and thereby to increase the internal validity of the study. Unfortunately, the results of rigorously controlled RCTs with strong internal validity are often not highly generalizable to the real-world setting of actual clinical practice. The aim of this paper is to briefly review key variables that influence treatment outcome of bipolar disorder in real-world clinical settings and to discuss strategies for optimizing treatment outcomes under ordinary conditions.