Objectives: To compare individual first-generation antipsychotics (FGAs) with individual second-generation antipsychotics (SGAs) in adults (18 to 64 years) with schizophrenia, schizophrenia-related psychoses, or bipolar disorder, with a focus on core illness symptoms, functional outcomes, health care system utilization, and adverse events.
Data Sources: We conducted comprehensive searches in 10 electronic databases up to July 2011. We hand-searched conference proceedings, clinical trials registers, and reference lists of relevant studies. We contacted content experts and authors of relevant studies.
Methods: Two reviewers independently conducted study selection, assessed methodological quality, extracted data, and graded the strength of evidence. We conducted a descriptive analysis and performed meta-analyses when appropriate.
Results: We included 113 studies of schizophrenia (22 comparisons) and 11 studies of bipolar disorder (6 comparisons), and 1 study included both. Trials (n = 123) had an unclear (63 percent) or high (37 percent) risk of bias. Cohort studies (n = 2) had good methodological quality.
Core illness symptoms (global ratings and total scores): For schizophrenia, clozapine was more efficacious than chlorpromazine based on the one reported scale. Results for haloperidol versus olanzapine were discordant, with olanzapine favored for one scale but no differences based on two other scales. Haloperidol was favored over quetiapine based on one of four scales reported. No differences were found for haloperidol versus aripiprazole, clozapine, risperidone, and ziprasidone.
For bipolar disorder, haloperidol was favored over ziprasidone on the one scale reported. No differences were observed for haloperidol versus aripiprazole, olanzapine, or risperidone.
Functional outcomes and health care system utilization: Evidence came primarily from single studies and showed no differences between groups.
Adverse events: No differences were found in mortality for chlorpromazine versus clozapine and haloperidol versus aripiprazole, or in metabolic syndrome for haloperidol versus olanzapine. The most frequently reported adverse events with significant differences were extrapyramidal symptoms; in most cases, the SGA had fewer extrapyramidal symptoms than haloperidol.
Other outcomes: For schizophrenia, few differences were found across comparisons and outcomes. No differences were observed in health-related quality of life. For bipolar disorder, there were few comparisons or differences.
Subgroups: The most common subgroups were race and treatment resistance. No notable differences were found compared with overall results.
Conclusion: Few differences of clinical importance for outcomes of effectiveness were found. Patient-important outcomes were rarely assessed. Data were sparse for the four key adverse events deemed a priori to be most clinically important.