A recent meta-analysis examined all blinded clinical trials that directly compared second-generation antipsychotics, including 76 trials with 13,558 participants. The authors used the Positive and Negative Syndromes Scale (PANSS) as their primary outcome measure. On the PANSS, several differences emerged: olanzapine was signifi- cantly superior to aripiprazole, quetiapine, risperidone, and ziprasidone; risperidone outperformed quetiapine and ziprasidone. Fewer studies reported data separately on positive and negative symptoms. For positive symptoms, olanzapine and risperidone again outperformed quetiapine and ziprasidone. When considering only negative symptoms, quetiapine was superior to clozapine; no other significant differences emerged. In terms of dropout from trials due to ineffica- cy, olanzapine again outperformed risperi- done, ziprasidone, and quetiapine while clozapine was superior to risperidone (Leucht S et al., Am J Psychiatry 2009; 166:152-163).

TCPR’s Take: Several statistically significant differences emerged between treatments, but they were small. For example, 16 patients would need to be treated with olanzapine rather than ziprasidone to avoid one dropout due to inefficacy. Other differences between treatments were of a similar magnitude; differences between olanzapine and risperidone were especially small. Only five studies included participants in their first episode of psychosis; no differences emerged in these studies. Treatment-resistant patients generally did not fare better on clozapine than comparator medications. However, clozapine patients on doses above 400 mg/day fared better than patients on risperidone. The authors stated they were surprised that clozapine did not consistently ourperform other antipsychotics, since other research has suggested that clozapine possesses superior efficacy relative to other atypicals for treatment-resistant patients, although this area of research remains controversial (McEvoy JP et al., Am J Psychiatry 2006;163:600-610).Pharmaceutical sponsorship was not related to outcomes. Differences in side effects and cost appear more substantial than differences in efficacy, so clinicians should carefully consider these factors when treating patients with schizophrenia.