Exploring the statistical robustness of high strength of evidence rating

To explore further findings of our recent Evidence-based Practice Center (EPC) project that indicated low predictive validity of strength of evidence (SOE) ratings. We will determine whether effect estimates that systematic reviewers grade as high SOE have adequate statistical robustness to be in line with the current definition of high SOE, namely “estimates of effects are close to the true effect.” Specific Aims: 1. Using TSA, we will assess a random sample of high SOE outcomes from Cochrane and AHRQ systematic reviews to determine the statistical robustness of these bodies of evidence. 2. When TSA results do not support high SOE grades, we will explore reasons for this discrepancy.