Geographic Variation in Medicare Spending and the Health, Functioning, and Behavioral Risk Factors of Older Americans
Embargo until
2020-08-01
Date
2016-03-29
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Johns Hopkins University
Abstract
Variation in healthcare delivery is ubiquitous in the United States. Two fundamental questions in health policy discussions today are these: what determinants explain the marked variation in healthcare spending across hospital referral regions in the United States and whether or not greater healthcare spending improves patient outcomes. This dissertation extends prior research by (1) examining the role that health behaviors and modifiable risk factors play in explaining differences in Medicare spending across regions and (2) evaluating how differences in spending affect beneficiaries’ physical, cognitive, and mental health and functioning.
Dissertation aims were addressed using data from the Health and Retirement Study (HRS), a nationally representative longitudinal survey of older Americans, and this was linked to Medicare claims and regional spending characteristics from The Dartmouth Atlas of Health Care. Medicare beneficiaries age 65 or older and enrolled in traditional (fee-for-service) Medicare comprised the study population. Price-adjusted Medicare spending was measured as spending for care in a variety of settings (inpatient, skilled nursing facilities, outpatient, physician office, home health, and hospice) as well as spending for durable medical equipment. Analytic methods included regression-based decomposition techniques and instrumental variables analysis.
In assessing the determinants of regional variation in spending, Medicare beneficiaries’ observable characteristics (e.g., sociodemographics, behavioral risk factors, health status factors) collectively explained 17% of regional differences in price-adjusted Medicare spending. Behavioral risk factors, specifically—smoking status, alcohol consumption, body mass index, and physical activity—explained 7% of the difference in spending between higher- and lower-spending regions.
In examining the effects of Medicare spending on health and functional outcomes among hospitalized Medicare beneficiaries (after adjusting for confounding due to health status), a 10% increase in price-adjusted, 1-year post-admission Medicare spending was associated with reductions in the probability of new limitations in instrumental activities of daily living (-1.96 percentage points; 95% confidence interval [CI], -3.88 to -0.03; P=0.05), new depressive symptoms (-2.31 percentage points; 95% CI, -4.04 to -0.59; P=0.009), and 1-year mortality (-2.02 percentage points; 95% CI, -3.57 to -0.46; P=0.01). There was no association between higher Medicare spending and self-rated health status, limitations in activities of daily living, pain, or cognitive functioning.
This dissertation provides policymakers with new information about the importance of behavioral risk factors as determinants of regional variation in Medicare spending and the impact of healthcare spending on multiple dimensions of health and functioning in Medicare beneficiaries.
Description
Keywords
Medicare spending, geographic variation, practice variation, healthcare costs