In Reply.—
Since OCD is generally characterized by only partial improvement with effective treatment, we agree with Tamimi and Mavissakalian that the choice of initial and subsequent pharmacotherapy represents a particularly important issue. However, we believe their letter is an attempt to "slice too small a pie into too many pieces" in that our study and their observations require further study and replication. We certainly agree, and have already suggested in our article, that comparative data from more patients are required before guidelines for preferred treatment choices can be promulgated.Tamimi and Mavissakalian express concern regarding our suggestion that clinicians consider performing a trial of fluoxetine in patients with OCD who do not respond to clomipramine, and vice versa. Instead, they suggest that our study, and another study by Jenike and coworkers1 that also compared clomipramine to fluoxetine in patients with OCD, indicated that nonresponders to clomipramine are not likely to