Toward a better understanding of the role of geography in intensity of end-of-life care: must we first come to an understanding of end-of-life care?

摘要

Beginning in the 1970s, Wennberg and colleagues1 have used health care claims (or administrative data) combined with mapping to forward a supplier-induced demand hypothesis. They argue that it is the providers, not the patients, who are fueling differences in the use of certain types of "high variation" health care between areas of the United States and even between similar cities, such as Boston and New Haven.2 Even before Scitovsky's landmark 1984 study,3 there was a recognition that the large amount of money spent providing care for people hi then" last 6 months of life signaled a mismatch between the amount and type of care desired and the care received. In this month's issue ofMedical Care, Barnato and colleagues4 present the results of a survey linking self-reported preferences for end-of-life care with hospital referral region spending in the last 6 months of life for Medicare beneficiaries. She and her colleagues surveyed 2515 elderly Medicare beneficiaries living in 290 of 306 Hospital Referral Regions (HRRs) and assessed then-preferences for aggressive end-of-life care. They did not find any association between preferences for aggressive end-of-life care and HRR-level measures of intensity for hospital and physician spending in the last 6 months of life. The article adds to the growing body of literature surrounding geographic variation in end-of-life care hi the United States and the debate about the role of local patterns of care on end-of-life care.