Dr. Leverich Replies

To the Editor: We thank Dr. Mattes for commenting on our article on the risk of switching during antidepressant treatment. Dr. Mattes is correct that there is some ambiguity in the clinical meaning of the ratio of the number of full switches to the number of subthreshold switches across different drugs. However, several factors suggest that the greater ratio of full to subthreshold switches with venlafaxine relative to the other antidepressants is a meaningful result.

The first factor is an important methodological one. If patients had one or more brief hypomanias and then progressed to having full duration or more severe episodes, only the latter, most severe episode was counted. Thus, the apparent decrease in minor or subthreshold switches appears to be related to the fact that more patients receiving venlafaxine progressed to full duration hypomanias or manias relative to the patients given the other drugs.

Also consistent with this interpretation are the findings, based on this same cohort, that were analyzed using conventional cross-sectional measures of manic severity as reported in a companion article (1) . If one used a two-point or greater increase in the Clinical Global Impression Scale for Bipolar Disorder mania severity rating as indicating a clinically meaningful switch, this criterion occurred in more patients receiving venlafaxine (p<0.01). Similarly, if one used the criterion of a Young Mania Rating Scale score of >13 as an indication of clinically meaningful hypomania or mania, this also occurred more frequently with venlafaxine (p=0.05), especially in rapid cyclers. Moreover, Vieta and colleagues (2) also found higher switch rates with venlafaxine than paroxetine in a randomized open study.

Thus, the graded Young Mania Rating Scale and Clinical Global Impression Scale for Bipolar Disorder manic severity ratings used by Post and colleagues (1) and the methodological clarification that only the most severe form of manic switch observed was counted suggest that there is an increased proclivity for full switches with venlafaxine relative to the other two drugs (particularly bupropion). On the basis of both analyses, we would conclude that greater caution is warranted (particularly in those patients with a prior history of four or more episodes in the year prior to study entry) in the use of antidepressant adjunctive treatment with venlafaxine than sertraline or bupropion in the treatment of depressed patients with bipolar disorder who are already receiving a mood stabilizer.