Bayesian methods of analysis for cluster randomized trials with binary outcome data

We explore the potential of Bayesian hierarchical modelling for the analysi s of cluster randomized trials with binary outcome data, and apply the meth ods to a trial randomized by general practice. An approximate relationship is derived between the intracluster correlation coefficient (ICC) and the b etween-cluster variance used in a hierarchical logistic regression model. B y constructing an informative prior for the ICC on the basis of available i nformation, we are thus able implicitly to specify an informative prior for the between-cluster variance. The approach also provides us with a credibl e interval for the ICC for binary outcome data. Several approaches to const ructing informative priors from empirical ICC values are described. We inve stigate the sensitivity of results to the prior specified and find that the estimate of intervention effect changes very little in this data set, whil e its interval estimate is more sensitive. The Bayesian approach allows us to assume distributions other than normality for the random effects used to model the clustering. This enables us to gain insight into the robustness of our parameter estimates to the classical normality assumption. In a mode l with a more complex variance structure, Bayesian methods can provide cred ible intervals for a difference between two variance components, in order f or example to investigate whether the effect of intervention varies across clusters. We compare our results with those obtained from classical estimat ion, discuss the relative merits of the Bayesian framework, and conclude th at the flexibility of the Bayesian approach offers some substantial advanta ges, although selection of prior distributions is not straightforward. Copy right (C) 2001 John Wiley & Sons, Ltd.