sicians. For example, for the average patient in Miami, Medicare spends about two and a half times the amount it spends for the average patient in Minneapolis (Wennberg et al., 1999), even after adjusting for age, illness severity, and comorbidities; in recent years, Medicare spent an average of almost $60,000 on New Jersey patients in the last 24 months of their lives, but only half that amount on similar patients in North Dakota (Wennberg et al., 2008). At the same time, effective preventive services, such as mammography or pneumonia vaccinations, are underutilized in both high- and low-cost geographic areas (Wennberg et al., 1999).

Ironically, a greater intensity of services does not necessarily mean that patients fare better. Sometimes, they fare worse. Mortality rates for patients with the same personal characteristics and the same disease are often higher in locales where more health care services are routinely provided (Wennberg et al., 2008).

Significant resources could be saved throughout the health system if the least efficient providers mimicked the practices of the most efficient (Antos and Rivlin, 2007). If all patients nationwide had the kind and intensity of care that patients receive in the least-intensive, most conservative settings (notably Mayo Clinic in Rochester, Minnesota, and Intermountain Healthcare in Salt Lake City), Medicare—and perhaps other—spending could be reduced by about 30 percent (Wennberg et al., 2002). There may always be patients who do benefit more from an intensive approach, but the costs of paying for extra care for these few would be more than balanced by reducing the intensity of services for the larger number who receive too much care (Wennberg et al., 2008).

Figure 4-1 presents state-level data showing what Medicare spends, on average, per beneficiary, compared to how the quality of care for beneficiaries is rated in that state. Each dot represents a state, and the figure clearly shows the absence of a relationship between spending and quality. If the two were related, low-spending states would be clustered in the lower left of the figure and higher-spending states would rise on the quality scale. Instead, beneficiaries in some states on the low end of the spending scale receive high-quality care, whereas beneficiaries in some states on the high end of the spending scale receive low-quality care. In fact, the state with the highest-quality care is at the low-cost end, with annual costs of about $6,000 per beneficiary, whereas the two states where care is most expensive (close to $9,000 per year) have among the lowest quality ratings.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement