To the Editor: I congratulate Drs. Volavka and colleagues on the publication of their landmark study comparing the efficacy of the atypical antipsychotics clozapine, risperidone, olanzapine, and haloperidol in patients with suboptimal response to conventional antipsychotics. The key feature of this study to a clinician like myself is that for the first time "apples and apples" were compared in a design that was hypothesis, not marketing, driven. All of the agents were compared with similar measures of efficacy and side effects. The medication doses used were similar to those used in clinical practice settings for patients with chronic illness in the schizophrenia spectrum. The commonly mentioned mean dose of 4.6 mg/day of risperidone does not work well in this group of patients. This study clearly tells us that in order to achieve therapeutic efficacy as determined by objective rating criteria, higher doses need to be used, as evident in the case of both risperidone and olanzapine, and that it can be done safely while managing neuroleptic-induced parkinsonism. This favors antipsychotic monotherapy in a subset of patients. In clinical practice, however, patients take combinations of antipsychotics in suboptimal doses without any support from the literature. Combination therapy is also costlier than monotherapy.

This study must have been difficult to conduct, and I am curious to know what extra steps the investigators took—if any—to maintain the blind with clozapine, because it causes extreme dry mouth, hypotension, sedation, and drooling. Data pertaining to neuroleptic-induced parkinsonism were reported: 30% of the risperidone group needed antiparkinsonian medication, while the percents were relatively low in the clozapine and olanzapine groups. Was there any differential in the rates of tardive dyskinesia even though the study duration was only 14 weeks? It would also be interesting to know if there was any difference in suicide attempts or completed suicides among patients in the various medication groups. Could the authors also comment on glucose or lipid abnormalities in the groups?

This study is an excellent example of a fruitful partnership between industry and federal funding, with scientific interests and patient care given their due.