To the Editor: The double-blind, randomized, placebo-controlled trial of Paul E. Keck, Jr., M.D., et al. (1) suggests that aripiprazole is effective and safe for acute bipolar episodes—mania and mixed. We consider that careful scrutiny of the methodology and the results of this study are worthwhile before integrating the study results into clinical practice.

The study incorporated patients whose mania was below 4 weeks’ duration and excluded those with prior nonresponse to clozapine. Although the authors did not use a duration criterion for mixed affective episode, considering the unique pharmacodynamic properties of aripiprazole, it is unclear why such an exclusion of patients with severe and refractory mania was considered. In this study (1), there seems to be less uniformity in assessing the severity of psychopathology. While manic symptoms were quantified with the 11-item Young Mania Rating Scale (2), the measurement of severity of depression was performed with a single item of the Clinical Global Impression scale (bipolar version) (3). Because about one-third of the study participants had a mixed affective episode at inclusion and because improvement in depression was an outcome measure, use of a more quantifiable scale to quantify depression, e.g., the Hamilton Depression Rating Scale, could have made efficacy analysis balanced and more meaningful.

Another issue of concern is assessment variations. Although the authors attempted to control the effect of intercentric assessment variations by loading study centers as a covariate in analysis of covariance, which we consider an indirect way of addressing interrater reliability, they failed to address the details of intracentric assessment. Considering that this is a multicentric study, such a description of intracentric assessment variations, if any, is important for interpretation of the results. In this study (1), use of analysis of covariance to control the effect of baseline psychopathology is not adequately justified (1), as there was no mention that baseline psychopathological scores differed significantly between groups. A further issue, under the safety analysis section, the authors could have provided the details (mean dose and pattern) of anticholinergic medication use.

The high attrition rate observed in both groups (79% in the placebo group and 58% in the aripiprazole group) was the main limitation of this study and hinted that the study findings were only preliminary evidence of aripiprazole’s antimanic property. Thus, further studies are needed to arrive at a robust conclusion regarding the benefits of aripiprazole in acute bipolar—mania and mixed—episodes.