Does greater overall health care intensity from the provision of “supply-sensitive” medical care result in better health outcomes?

In the case of preference-sensitive care, the significance of the HEALT H AFFAIR S – We b Exclusiv eMEDICAR E REFOR M EXHIBIT 3 Comparison Of Medicare Spending, Supply-Sensitive Care, Preference-Sensitive Care, And Effective Care For Orange County, Miami, Minneapolis, And Portland Hospitai Referrai Regions, 1995-199 6 Ratio to Minneapolis region • Orange County (CA) 6.0 Miami (FL) • I Minneapolis (iVIN) • Portland (OR) Medicare spending SOURCE: Dartmouth Atlas of Health Care, 1995-96 database. NOTE: Rates are given as ratio to Minneapoils hospitai referrai region (vaiued as 1.0). ^ Care provided per decedent in the iast six months of iife. *> See Exhibit 2 for definitions. variation in use rates cannot be strictly interpreted from the point of view of the patients’ welfare, since it is not clear whether patients actually had much of a say in determining which treatment they received. Chnical studies of shared decision-making programs designed to inform patients about the treatment options available for low-back pain, prostatic hyperplasia, and stable angina do, however, suggest that the amount of surgery now provided in many regions exceeds what an informed Medicare population would demand.”’ Does greater overall health care intensity from the provision of “supply-sensitive” medical care result in better health outcomes? To address this question, we have evaluated the natural experiments afforded by the variations in care intensity among regions. Studies at the population level indicate no net advantage in terms of life expectancy for Medicare enrollees living in regions with more hospital resources (and hospitahzations) and greater care intensity as measured by more aggressive treatment patterns during the last six months of life.’^ Longitudinal (cohort) studies of patients with similar diseases (such as hip fracture) who have been followed for a number of years also show that patients living in high-careintensity regions gain no survival advantage over those in lowintensity regions.” HEALT H AFFAIR S – J 3 Februar y 200 2MEDICAR E The major limitation of these studies is the possibility that beneficiaries in high-spending regions could achieve gains in their quality of life. Several lines of research provide at least suggestive evidence that quality of life in high-intensity regions may not be better than in low-intensity regions. First, case-mix-acijusted longitudinal studies of Medicare beneficiaries found that those residing in highintensity regions achieved no gain in relief from angina or improvement in function.^” Second, two randomized trials testing the impact of greater medical care intensity for pal:ients with chronic disease found no benefit in terms of functional status and quahty of life.^’ Third, evidence from the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment (SUPPORT) study suggests a poor match between patients’ preferences and how patients with severe chronic illness are actually treated. Patients who stated that they would prefer an out-of-hospital death were no less likely to die in a hospital than were patients who expressed a preference for an in-hospital death. What did matter was local hospital capacity: The overall supply of hospital resources in the region effectively predicted whether the patient died in a hospital.^^ Be- —^^^•^•H cause most elderly people express a preference for a less intensive W104 MEDICARE approach to care as death approaches, greater intensity could lead to REFORM poorer quality of care among this group.