A comparison of national guidelines for network meta-analysis

Within technology appraisals, it is necessary to compare the complete set of treatments that may be used in the patient group under consideration. Randomized controlled trials are a key source of evidence for these comparisons. The techniques of network meta-analysis allow the networks of trial evidence to be evaluated to obtain estimates of comparative efficacy between sets of treatments. These techniques may be the only source of estimates of comparative effectiveness if trials directly comparing the treatments of interest have not been conducted, and may provide useful additional evidence if both direct and indirect comparisons exist.

We examined both published and draft guidelines from reimbursement and health technology appraisal bodies, and considered their recommendations using appropriate methodology for the conduct of indirect comparisons and the assessments of their validity.

Guidelines from 33 countries were reviewed. Of these, guidelines from 9 countries-Australia, Belgium, Canada, France, Germany, Scotland, Spain, South Africa, and the United Kingdom (England and Wales)-included detailed recommendations on the conduct of network meta-analysis. The recommendations were summarized.

No two recommendations from the multiple national guidelines are mutually exclusive. It is possible to perform one network meta-analysis for submission to multiple national jurisdictions.