Meta-analysis of drop-out rates in randomised clinical trials, comparing typical and atypical antipsychotics in the treatment of schizophrenia.

OBJECTIVE: To assess antipsychotic medication in the treatment of schizophrenia, based on trial drop-out rates. METHOD: The studies included were randomised controlled trials that compared any of the four clinically best-established atypical antipsychotics (quetiapine, olanzapine, risperidone or clozapine) against either of two typical antipsychotics regarded as the gold standard (haloperidol or chlorpromazine). RESULTS: Meta-analysis indicated less risk of all-cause patient withdrawal from atypical medication trials where dosage was flexible, in both the short, relative risk (RR) 0.70 (95% CI 0.64-0.76), P<0.00001, and long term, RR 0.72 (0.65-0.80), P<0.00001. Similar results were observed for withdrawal due to adverse events, RR: 0.54 (0.41-0.72), P<0.0001. Nevertheless, the favourable effects of atypical medication disappeared in trials relying on fixed dosage. CONCLUSIONS: We detected a significant positive effect in terms of the outcome of treatment discontinuation for atypical versus typical medication, though only where the use of flexible rather than fixed doses (closer to an experimental control situation) was possible.