Objectives: Treatment continuation, as measured by time to all-cause treatment discontinuation, is a broad measure of overall treatment effectiveness. This integrated analysis compared the likelihood of discontinuation from olanzapine treatment versus other antipsychotics among patients with schizophrenia.
Methods: Clinical trials of all sponsors were included if they met the following criteria: double-blind, randomized, comparative; duration of 12 weeks or longer; no mandatory discontinuation before 12 weeks; and schizophrenia-spectrum disorders; 20 patients or more per treatment. Weighted mean hazard ratios and 95% confidence intervals were calculated from discontinuation time. Meta-analyses were performed for the following comparators that had at least 2 studies: haloperidol (5 studies), risperidone (5 studies), ziprasidone (2 studies), clozapine (3 studies), and perphenazine (2 studies) (13 studies in total; 3 included more than 1 comparator). Only 1 eligible published study was found for fluphenazine, amisulpride, and quetiapine; therefore, meta-analyses could not be performed for these comparators.
Results: Significantly (P < 0.05) greater likelihood of discontinuation relative to olanzapine treatment (hazard ratio [95% confidence interval]) was observed for haloperidol (1.4 [1.2-1.7]), risperidone (1.3 [1.1-1.6]), ziprasidone (1.6 [1.4-2.0]), and quetiapine (1.4 [1.1-1.9]), but not clozapine (1.2 [0.9-1.6]), fluphenazine (1.8 [0.8-4.3]), perphenazine (1.3 [0.7-2.1]), or amisulpride (1.1 [0.8-1.6]).
Conclusions: These data suggest that patients with schizophrenia and related disorders may continue olanzapine treatment longer than haloperidol, risperidone, ziprasidone, or quetiapine treatment.