Bayesian inference for hospital quality in a selection model

John Geweke, Gautam Gowrisankaran, Robert J. Town

Research output: Contribution to journalArticle

80 Citations (Scopus)

Abstract

This paper develops new econometric methods to infer hospital quality in a model with discrete dependent variables and nonrandom 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 the highest quality. There is strong evidence of dependence between the unobserved severity of illness and the assignment of patients to hospitals, whereby patients with a high unobserved severity of illness are disproportionately admitted to high quality hospitals. Consequently a conventional probit model leads to inferences about quality that are markedly different from those in this study's selection model.

Original languageEnglish (US)
Pages (from-to)1215-1238
Number of pages24
JournalEconometrica
Volume71
Issue number4
StatePublished - 2003
Externally publishedYes

Fingerprint

Selection Model
Bayesian inference
illness
Selection model
Hospital quality
Probit Model
Mortality Rate
Posterior Probability
Markov Chain Monte Carlo
Econometrics
econometrics
Simulator
Assignment
mortality

Keywords

  • Bayesian inference
  • Hospital quality
  • MCMC
  • Medicare
  • Mortality
  • Pneumonia
  • Simultaneous equations

ASJC Scopus subject areas

  • Mathematics (miscellaneous)
  • Statistics and Probability
  • Economics and Econometrics
  • Social Sciences (miscellaneous)

Cite this

Bayesian inference for hospital quality in a selection model. / Geweke, John; Gowrisankaran, Gautam; Town, Robert J.

In: Econometrica, Vol. 71, No. 4, 2003, p. 1215-1238.

Research output: Contribution to journalArticle

Geweke, J, Gowrisankaran, G & Town, RJ 2003, 'Bayesian inference for hospital quality in a selection model', Econometrica, vol. 71, no. 4, pp. 1215-1238.
Geweke, John ; Gowrisankaran, Gautam ; Town, Robert J. / Bayesian inference for hospital quality in a selection model. In: Econometrica. 2003 ; Vol. 71, No. 4. pp. 1215-1238.
@article{8ba0a65d46b649a9b58972c6c0d288ef,
title = "Bayesian inference for hospital quality in a selection model",
abstract = "This paper develops new econometric methods to infer hospital quality in a model with discrete dependent variables and nonrandom 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 the highest quality. There is strong evidence of dependence between the unobserved severity of illness and the assignment of patients to hospitals, whereby patients with a high unobserved severity of illness are disproportionately admitted to high quality hospitals. Consequently a conventional probit model leads to inferences about quality that are markedly different from those in this study's selection model.",
keywords = "Bayesian inference, Hospital quality, MCMC, Medicare, Mortality, Pneumonia, Simultaneous equations",
author = "John Geweke and Gautam Gowrisankaran and Town, {Robert J.}",
year = "2003",
language = "English (US)",
volume = "71",
pages = "1215--1238",
journal = "Econometrica",
issn = "0012-9682",
publisher = "Wiley-Blackwell",
number = "4",

}

TY - JOUR

T1 - Bayesian inference for hospital quality in a selection model

AU - Geweke, John

AU - Gowrisankaran, Gautam

AU - Town, Robert J.

PY - 2003

Y1 - 2003

N2 - This paper develops new econometric methods to infer hospital quality in a model with discrete dependent variables and nonrandom 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 the highest quality. There is strong evidence of dependence between the unobserved severity of illness and the assignment of patients to hospitals, whereby patients with a high unobserved severity of illness are disproportionately admitted to high quality hospitals. Consequently a conventional probit model leads to inferences about quality that are markedly different from those in this study's selection model.

AB - This paper develops new econometric methods to infer hospital quality in a model with discrete dependent variables and nonrandom 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 the highest quality. There is strong evidence of dependence between the unobserved severity of illness and the assignment of patients to hospitals, whereby patients with a high unobserved severity of illness are disproportionately admitted to high quality hospitals. Consequently a conventional probit model leads to inferences about quality that are markedly different from those in this study's selection model.

KW - Bayesian inference

KW - Hospital quality

KW - MCMC

KW - Medicare

KW - Mortality

KW - Pneumonia

KW - Simultaneous equations

UR - http://www.scopus.com/inward/record.url?scp=0141606754&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0141606754&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0141606754

VL - 71

SP - 1215

EP - 1238

JO - Econometrica

JF - Econometrica

SN - 0012-9682

IS - 4

ER -