Interpreting Posterior Relative Risk Estimates in Disease-Mapping Studies