Estimating hospital quality based on performance measure and hospital acquired infection data: A Bayesian bivariate approach

Hospital quality data consists of dichotomous process measurements and count outcome measures. However, current statistical approaches to evaluating hospital quality often utilize one of these measurements. Moreover, when data are collected from multiple units within a hospital, the within-hospital correlation is typically ignored. To address this, we introduce a novel methodology for determining hospital quality that incorporates both outcomes within a Bayesian framework. We illustrate our approach using performance measures obtained from the Joint Commission and hospital acquired infection measures directly from hospitals. Using our approach, we differentiate high and low performing hospitals with a high degree of certainty.