How to Use an Article Reporting a Multiple Treatment Comparison Meta-Analysis
What is the best approach to critically evaluate studies using multiple treatment comparison (MTC) meta-analyses?
This manuscript summarizes a systematic approach to critical interpretation of studies using MTC meta-analyses, and describes common biases that readers should consider when reviewing these studies.
MTC meta-analyses can be used to estimate the effect of pairwise comparisons between interventions when inadequate head-to-head randomized, controlled trials are available. This approach can combine both direct comparisons (A vs. B) and indirect comparisons (for example, A vs. placebo and B vs. placebo) to provide estimates of the comparative effect between interventions. When critically appraising these studies, the first step is to determine whether the results are valid. For this, one needs to ensure that the clinical question is logical and that all relevant studies are included, and assess for the presence of bias in the studies is utilized. Common sources of bias in these studies include publication bias and selective outcome reporting bias, and some comparisons may be affected by these biases more than others. The second step is to review the results, and to assess whether these are consistent between studies, and between direct and indirect comparisons. Further, the overall effect size must be considered in context of the potential biases that may be present. The final step is to decide whether these findings are clinically relevant. This includes verifying that all important interventions and outcomes were assessed, and evaluating the overall robustness of the study and its main limitations.
A systematic approach should be used to determine the validity and clinical applicability of results from MTC meta-analyses.
Medical literature frequently lacks an adequate number of direct head-to-head comparisons between alternate interventions, and for some clinical questions, many of the available studies provide only indirect comparisons. MTC meta-analyses may be a useful means to combine both direct and indirect comparisons of interventions, but these analyses have inherent limitations and should be critically assessed in a systematic manner.
Keywords: Publication Bias
< Back to Listings