FOR-PROFIT HOSPITAL OWNERSHIP AND MEDICARE SPENDING

Dartmouth College, Hanover NH

Investigators

Abstract

As increasing numbers of hospitals are converting from not-for-profit to for- profit ownership, the proper role of for-profit enterprise in health care continues to be debated. Some feel that the profit motive of investor-owned businesses threatens to inflate costs and decrease quality of health care delivery. Our previous work has demonstrated an association between for- profit hospital ownership and higher per capita Medicare spending, but the exact mechanisms by which increased spending occurs remain undefined. The objectives of this project are 1) to further examine the relationship between hospital ownership and Medicare spending in all segments of health care and 2) to describe the specific patterns of hospital spending and utilization that contribute to spending differences. Analyses will be population-based, with the unit of analysis being the hospital service area (HSA) as previously defined in the Dartmouth Atlas of Healthcare. Using American Hospital Association data, each HSA will be categorized as for-profit (all beds in the area are in for-profit hospitals), not-for-profit (all beds in the area are in not-for-profit hospitals) or mixed for each year between 1992 and 1998. To accomplish our first goal, data from the Continuous Medicare History Sample will be used to calculate per capita Medicare spending rates in each area. Total spending will also be broken down by type of service: inpatient hospital, outpatient hospital, home health services, skilled nursing facilities, and physician services. To accomplish our second goal, 100% of all Medicare hospital discharges will be collected for 18 states that have experienced hospital conversions within the study period. We will compare volume, intensity, and price of hospital services by HSA ownership, and HSAs with stable ownership will be compared to those in which ownership converted within the study period. For each analysis we will use multivariate regression models to control for patient demographics and other HSA characteristics known to influence spending and health care utilization.