Grant: Fellowship - PA: Clinical Research

Background: Systematic reviews can be prone to different types of heterogeneity. Variability in the participants, interventions and outcome characteristics may be termed clinical heterogeneity; variability in the trial design and quality is typically termed methodological heterogeneity; variability in summary treatment effects between trials can be termed statistical heterogeneity. While guidance exists on assessment of methodological and statistical heterogeneity, little attention has been given to clinical heterogeneity. Therefore, we propose to develop a set of method guidelines for assessing clinical heterogeneity in systematic reviews of randomized clinical trials. Also, we propose to assess a selection of published systematic reviews with the completed guidelines. Methods: Guideline development will consist of a modified Delphi process of three phases: 1. Pre-meeting item generation, 2. Face to face consensus meeting, and 3. Post-meeting feedback. During phase 1 identified participants will be sent articles collected during a literature search, then contacted by telephone and asked to suggest methods necessary for assessing clinical heterogeneity in systematic reviews of randomized clinical trials. Participants will be asked to consider methods based upon empirical evidence and secondarily common sense reasoning. In the second phase we will hold a consensus meeting during which each method will be discussed and debated for relevance to assessing clinical heterogeneity. Final method suggestions and operational definitions will be developed and modified based upon discussions and group consensus. In phase 3 the final guidance document will completed and circulated to meeting participants for feedback. Finally, we will sample 100 systematic reviews and assess them with the guidelines to identify deficiencies.