Estimating the quality of care in hospitals using instrumental variables

Abstract

Mortality rates are a widely used measure of hospital quality. A central problem with this measure is selection bias: simply put, severely ill patients may choose high quality hospitals. We control for severity of illness with an instrumental variables (IV) framework using geographic location data. We use IV to examine the quality of pneumonia care in Southern California from 1989 to 1994. We find that the IV quality estimates are markedly different from traditional GLS estimates, and that IV reveals different determinants of quality. Econometric tests suggest that the IV model is appropriately specified, that the GLS model is inconsistent.

Spatial disparities in hospital performance

“…15 Regions at extremes are the same as when studying the raw data: Alsace (at the German border) usually exhibits the highest survival function and Languedoc-Roussillon (in the South-East) the lowest. Graph 5 represents the survival functions (as well as their confidence intervals) for these two extremes and for Ile-de-France (The Paris region).…”

The Relationship between Quality and Hospital Case Volume – An Empirical Examination with German Data

“…Furthermore, hospital choice models sensibly assume that there is a correlation between patients' severity of disease and hospital choice (Gowrisankaran and Town, 1999). As we have the zip codes of all our patients and the exact address of the hospital we computed travel times for all our patients (shown in the Appendix).…”

Schooling, Child Labor, and the Returns to Healthcare in Tanzania

“…Thus, comparing outcomes across individuals who used different healthcare options will lead to a biased estimate of the impact of using better healthcare. Moreover, fixed effects estimators cannot adequately address this problem because they do not control for unobserved severity, which varies by illness episode (that is, within multiple observations of the same individual), and which other studies have shown is the most salient bias for acute shocks (Gowrisankaran and Town 1999;Cutler, Huckman, and Landrum 2004).…”