Abstract
Treatment Discontinuation Following Randomization to Open-Label Olanzapine, Risperidone or Typical Antipsychotics During a One-Year Treatment for Schizophrenia
Author(s): Haya Ascher-Svanum , Allen W. Nyhuis , Douglas E. Faries ,Lenore Heiler , Bruce J. KinonObjective: To compare olanzapine, risperidone and typical antipsychotics on medication discontinuation for any cause during one-year, open-label treatment for schizophrenia. Methods: This post hoc analysis used data from a one-year, randomized, open-label, multi-site, cost-effectiveness study of schizophrenia conducted between May 1998 and September 2002. If clinically warranted, patients could switch antipsychotics. Patients randomized to olanzapine (n=222), risperidone (n=217) or typical antipsychotic of physician’s choice (n=209) were compared on average time to allcause medication discontinuation and survival rates over the one-year study. Results: The average time to all-cause discontinuation was longer for olanzapine than for risperidone (p<.001), typical antipsychotics (p<.001), and versus perphenazine, a medium potency typical antipsychotic (p=.002). Treatment with risperidone was longer than with typicals (p<.01), but not significantly different from perphenazine. One-year survival rate was higher for olanzapine therapy (55%) compared to risperidone (47%, p=.006), typical antipsychotics (32%, p≤.001) and perphenazine (31%, p<.001). Survival rate for risperidone-treated patients was higher than on typicals (p<.001), but not significantly different from perphenazine. Patients randomly assigned to continue their baseline medication (“stayers”) had significantly longer times until discontinuation than did those assigned to switch antipsychotics, but treatment group differences were essentially unchanged when analyses excluded the “stayers.” Conclusions: In the long-term, open-label treatment of patients with schizophrenia, antipsychotics appear to significantly differ on time to all-cause medication discontinuation, a measure considered a proxy index of a medication’s effectiveness. Findings were essentially unchanged when treatment group comparisons took into account whether the medications being compared were newly initiated