Bayesian Inference for Hospital Quality in a Selection Model, Issue 8497
National Bureau of Economic Research, 2001 - Bayesian statistical decision theory - 43 pages
This paper develops new econometric methods to infer hospital quality in a model with discrete dependent variables and non-random selection. Mortality rates in patient discharge records are widely used to infer hospital quality. However, hospital admission is not random and some hospitals may attract patients with greater unobserved severity of illness than others. In this situation the assumption of random admission leads to spurious inference about hospital quality. This study controls for hospital selection using a model in which distance between the patient's residence and alternative hospitals are key exogenous variables. Bayesian inference in this model is feasible using a Markov chain Monte Carlo posterior simulator, and attaches posterior probabilities to quality comparisons between individual hospitals and groups of hospitals. The study uses data on 74,848 Medicare patients admitted to 114 hospitals in Los Angeles County from 1989 through 1992 with a diagnosis of pneumonia. It finds the smallest and largest hospitals to be of high quality and public hospitals to be of low quality. There is strong evidence of dependence between the unobserved severity of illness and the assignment of patients to hospitals. Consequently a conventional probit model leads to inferences about quality markedly different than those in this study's selection model.
What people are saying - Write a review
We haven't found any reviews in the usual places.
114 hospitals admitted to hospital alternative priors Angeles County base model Bayesian Inference beds CHARTER COMMUNITY choice model coefficients column COMMUNITY HOSPITAL component computation conventional probit model covariates demographic differences discharge records Disease stage 3.8 DOCTORS HOSPITAL Econometrics function Gaussian Geweke Gibbs sampling algorithm Gowrisankaran group hospital quality group quality probits Health hospital admission hospital choice HOSPITAL MEDICAL CENTER Hospital name hospital quality probits hyperparameters income independent indicator instrumental variables iterations KAISER FOUNDATION HOSPITAL latent variables Markov chain matrix McClellan means and standard MEMORIAL HOSPITAL mortality equation mortality probit equation mortality rates multinomial probit model NBER Working Paper non-random Panel pneumonia posterior distribution posterior means posterior moments posterior probability posterior standard deviation prior distribution prior standard deviations prior variant Private For-profit provides quartile SANTA MARTA HOSPITAL severity correlations severity of illness SHERMAN OAKS subscription Table TEMPLE COMMUNITY HOSPITAL unobserved disease severity unobserved severity variance variation in hospital zip code