The Choice of a Noninformative Prior on Between-Study Variance Strongly Affects Predictions of Future Treatment Effect

Purpose. Bayesian random-effects meta-analyses require the analyst to specify the prior distribution for between-study variance of the treatment effect. We assessed the sensitivity of prediction and other outputs of the meta-analysis to the choice of this prior. Methods. We reana-lyzed 7 published meta-analyses (5–14 trials) with rare (event rates \5%), moderate (15%–50%), and frequent binary outcomes (.50%). We examined 10 noninforma-tive priors: inverse gamma on between-study variance (t2), 2 uniforms on each of the between-study standard deviation (t) and t2, uniform shrinkage on t2, DuMouchel shrinkage on t, half-normal on t2, and half-normal priors on t with large and small variances. For each analysis, we calculated the posterior distributions for t, the population treatment effect in current studies, and the predicted treat