Haloperidol dose when used as active comparator in randomized controlled trials with atypical antipsychotics in schizophrenia: comparison with officially recommended doses.

OBJECTIVE: To determine the doses of haloperidol as a comparator drug in randomized controlled trials (RCTs) with atypical antipsychotics in patients with schizophrenia and to compare these doses with the officially recommended doses for haloperidol in the United States and the United Kingdom.

DATA SOURCES: We searched for RCTs conducted and published in English in full before January 2005 in which atypical antipsychotics were compared with haloperidol for the treatment of schizophrenia. We searched for Cochrane Reviews in which 1 of the following atypical antipsychotics was evaluated for the treatment of patients with schizophrenia, schizophreniform psychosis, or other primary psychosis: amisulpride, aripiprazole, olanzapine, quetiapine, risperidone, sertindole, and ziprasidone. For the gap between the end point of inclusion of the studies in the Cochrane Reviews and January 2005, we electronically searched the Cochrane Central Register of Controlled Trials for any further RCTs in which atypical antipsychotics were compared with haloperidol for the same indication. Search terms used were haloperidol and schizophren and haloperidol and psychotic, as well as the names of the selected atypical antipsychotics for the years that were not covered by the Cochrane Reviews. For each study, the required dose and mean dose of haloperidol were compared with officially recommended doses of haloperidol in U.S. (Food and Drug Administration) and U.K. (British National Formulary) guidelines.

DATA SYNTHESIS: In all of the included studies (N = 49), the midpoints of the required doses were above the midpoint of the official recommended doses in the United States and United Kingdom for moderately ill patients. In 94% (U.S.) and 80% (U.K.) of the studies, they were above the upper border of the recommended doses. Compared with recommended doses for severely ill patients in both the United Kingdom and United States (range, 6-15 mg daily), in 17 studies (35%) the mean actual used dose was above the upper dose border for severely ill patients (15 mg daily).

CONCLUSIONS: Nearly all randomized clinical trials used haloperidol in doses that were higher than the official recommended doses for moderately ill or even severely ill patients. Therefore, it is probable that the results of the RCTs were affected by the high dose of haloperidol, hampering the interpretation of the effects of atypical anti-psychotics in their comparison with haloperidol.