Insurance type and choice of hospital for coronary artery bypass graft surgery.

Abstract

OBJECTIVE: To examine the effect of insurance type on the relationship between hospital attributes and patient flows, with particular attention to whether HMO enrollees are more or less likely than other patients to receive care at high-quality hospitals and whether HMO enrollees travel farther to receive care. DATA SOURCES/STUDY SETTING: Data on patient flows, taken from discharge abstracts compiled by the California Office of Statewide Health Planning and Development. Our sample consists of patients undergoing coronary artery bypass graft surgery (CABG) in 1991 who resided in three California markets. Only patients under the age of 65 and insured by HMOs, Blue Cross/Blue Shield, or other commercial insurance were included. Hospital quality is based on hospital-specific measures of excess mortality from CABG. Other hospital attributes were taken from American Hospital Association survey data. STUDY DESIGN: Conditional-choice models were used to estimate the probability that patients would receive care at any given hospital as a function of their insurance type and the hospital's attributes. PRINCIPAL FINDINGS: Patients were more likely to receive care from hospitals closer to their residence. However, HMO patient flows were less sensitive to proximity. In general, the likelihood that an HMO enrollee received care at a given hospital was positively related to hospital quality. Moreover, quality had a greater effect on patient flows for HMO enrollees than for non-HMO enrollees. However, the evidence suggests that the effect of quality on patient flows is neither uniform across markets nor across HMOs. CONCLUSIONS: HMOs do not appear to direct patients to low-quality hospitals. However, heterogeneity among HMOs and across markets suggests that buyers must recognize that choosing an HMO involves greater scrutiny than simply picking a plan labeled "HMO."

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 80 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.