Empirical Bayes methods for estimating hospital‐specific mortality rates

Abstract

We present alternative methods for estimating hospital-level mortality rates to those used by the Health Care Finance Administration for Medicare patients. We use an empirical Bayes model to represent the different sources of variation in observed hospital-specific mortality rates and we use a logistic regression model to adjust for severity differences (in patient mix) across hospitals. In addition to providing a principled derivation of a standard error for the commonly used estimator, our fully model-based formulation produces much more accurate estimates and resolves the severe problem of multiple comparisons that arises when extreme estimates are used to identify exceptional hospitals. We estimate models for each of four disease conditions using the national Medicare mortality data base which does not contain patient severity descriptors, and mortality data from national samples which do include patient severity descriptors. We find substantial between-hospital variation in the unadjusted death rates from the national data base. Mortality rates differ substantially with patient severity in our models, but the sample sizes are too small to yield reliable estimates of the between-hospital variation in adjusted mortality rates.

Statistical and Substantive Inferences in Public Health: Issues in the Application of Multilevel Models

“…Indeed, much of the literature on risk adjustment in performance evaluation focuses on the proper specification of the risk-adjustment model (e.g., 6, 99, 100). A second difficulty is that, even with the increased reliability of empirical Bayes estimators, the resulting indicators of provider performance may still be too unstable to permit a useful comparison of individual providers (53,65,145).…”

Medicare Claims Can Be Used to Identify US Hospitals With Higher Rates of Surgical Site Infection Following Vascular Surgery

“…25,27,32,33 The rankings used empirical Bayes estimators (predicted random effects), which drew hospitals with low procedure volume toward the mean, due to their greater variability. [34][35][36] On the basis of the risk-adjusted rankings, hospitals were grouped into deciles.…”

Penalized regression with individual deviance effects

“…We note, however, that the estimated deviance effects and their standard errors allow the implementation of probabilistic ranking procedures (van Houwelingen et al 2004;Spiegelhalter 1999;Goldstein and Spiegelhalter 1996;Thomas et al 1994). We will report on this elsewhere.…”