Limited treatment responses in schizophrenia prompted the testing of combining an antipsychotic drug treatment with a second psychotropic medication. A comprehensive evaluation of the efficacy of multiple medication combinations is missing.Objective
To summarize and compare the meta-analytically determined efficacy of pharmacologic combination strategies of antipsychotic drugs in adults with schizophrenia.Data Sources
Systematic search of PubMed and PsycInfo until May 13, 2016.Study Selection
Meta-analyses of randomized clinical trials comparing the efficacy of antipsychotic drugs combined with other antipsychotic or nonantipsychotic medications vs placebos or antipsychotic monotherapy among adults with schizophrenia.Data Extraction and Synthesis
Independent reviewers extracted the data and assessed the quality of the methods of the included meta-analyses using A Measurement Tool to Assess Systematic Reviews (AMSTAR), adding 6 new items to rate their quality. Effect sizes, expressed as standardized mean difference /Hedges g or risk ratio, were compared separately for combinations with any antipsychotic drug and for combinations with clozapine.Main Outcomes and Measures
The primary outcome was total symptom reduction. Secondary outcomes included positive and negative symptoms, treatment recommendations by authors, study-defined inefficacies, cognitive and depressive symptoms, discontinuation of treatment because of any cause, and inefficacies or intolerabilities.Results
Of 3397 publications, 29 meta-analyses testing 42 combination strategies in 381 individual trials and among 19 833 participants were included. For total symptom reductions, 32 strategies that augmented any antipsychotic drug and 5 strategies that augmented clozapine were examined. Fourteen combination treatments outperformed controls (standard mean difference/Hedges g, −1.27 [95% CI, −2.35 to −0.19] to −0.23 [95% CI, −0.44 to −0.02]; P = .05). No combination strategies with clozapine outperformed controls. The quality of the methods of the meta-analyses was generally high (mean score, 9 of a maximum score of 11) but the quality of the meta-analyzed studies was low (mean score, 2.8 of a maximum score of 8). Treatment recommendations correlated with the effect size (correlation coefficient, 0.22; 95% CI, 0.35-0.10; P < .001), yet effect sizes were inversely correlated with study quality (correlation coefficient, −0.06; 95% CI, 0.01 to −0.12; P = .02).Conclusions and Relevance
Meta-analyses of 21 interventions fully or partially recommended their use, with recommendations being positively correlated with the effect sizes of the pooled intervention. However, the effect sizes were inversely correlated with meta-analyzed study quality, reducing confidence in these recommendations. Higher-quality trials and patient-based meta-analyses are needed to determine whether subpopulations might benefit from combination treatment, as no single strategy can be recommended for patients with schizophrenia based on the current meta-analytic literature.