RESUMEN
BACKGROUND AND HYPOTHESIS: Meta-analyses have shown that the majority of patients with schizophrenia who have not improved after 2 weeks of treatment with an antipsychotic drug are unlikely to fully respond later. We hypothesized that switching to another antipsychotic with a different receptor binding profile is an effective strategy in such a situation. STUDY DESIGN: In total, 327 inpatients with an acute exacerbation of schizophrenia were randomized to double-blind treatment with either olanzapine (5-20 mg/day) or amisulpride (200-800 mg/day). Those patients who had not reached at least 25% Positive-and-Negative-Syndrome-Scale (PANSS) total score reduction from baseline after 2 weeks (the "non-improvers") were rerandomized double-blind to either staying on the same compound ("stayers") or to switching to the other antipsychotic ("switchers") for another 6 weeks. The primary outcome was the difference in the number of patients in symptomatic remission between the combined "switchers" and the "stayers" after 8 weeks of treatment, analyzed by logistic regression. STUDY RESULTS: A total of 142 nonimprovers were rerandomized at week two. 25 (45.5 %) of the 'stayers' compared to 41 (68.3 %) of the "switchers" reached remission at endpoint (p = .006). Differences in secondary efficacy outcomes were not significant, except for the PANSS negative subscore and the Clinical-Global-Impression-Scale. "Switchers" and "stayers" did not differ in safety outcomes. CONCLUSIONS: Switching "non-improvers" from amisulpride to olanzapine or vice-versa increased remission rates and was safe. The superiority in the primary outcome was, however, not paralleled by significant differences in most secondary efficacy outcomes and the effect was only apparent at the last visit making replications of longer duration necessary.