Direct versus indirect comparisons: a summary of the evidence.

OBJECTIVES

The aim of this study was to provide a comprehensive summary and interpretation of the current evidence on the use and validity of statistical methods to conduct indirect comparisons of treatment effects.

METHODS

A narrative review was conducted.

RESULTS

Well-conducted methodological studies provide good evidence that adjusted indirect comparisons can lead to results similar to those from direct comparisons. The internal validity of several statistical methods to conduct indirect comparisons, therefore, has been established. Meta-regression, logistic regression, or adjusted indirect comparisons should be the methods of first choice. Unadjusted indirect comparisons are always unacceptable. Deciding whether to combine direct and indirect evidence will be a matter of informed judgment based on the similarities and dissimilarities of populations and interventions. Unverifiable assumptions with respect to the similarity of compared studies and low power are serious limitations of indirect comparisons.

CONCLUSIONS

In the absence of sufficient head-to-head evidence, adjusted indirect comparisons can be viewed as additional analytical tools to determine the comparative efficacy and effectiveness of competing interventions. Researchers who use indirect comparisons need to keep the limitations in mind.