|
FROM THE FIELDRising Medicare Costs: The Authors Respond
WE READ WITH INTEREST Dana Goldman and Neeraj Sood’s reaction to our paper on the rise in Medicare spending.1 We agree with the general thrust of their remarks, and we thank them for their careful reading of our paper. They raise several points that are worth highlighting. Our analysis examines trends in real per capita Medicare spending at two points in time: 1987 and 2002. Projecting future spending trends could build off our results, but such projections would have to include estimates of trends in lifetime spending. We agree that the implications for long-term Medicare spending trends depend critically on beneficiaries’ lifetime spending projections. One interesting implication of our results is that a key finding of the work by James Lubitz and colleagues might not hold true in the future.2 If the clinical thresholds for defining disease continue to change and include a larger share of the population and if rates of treatment conditional on these definitions rise, the baseline level of spending among Medicare beneficiaries will rise year in and year out. This could result in higher lifetime spending even among relatively "healthy" Medicare beneficiaries relative to those with more serious medical conditions (who cost more per year but die earlier). A final comment concerns the puzzle Goldman and Sood mention. Recent work has found that disability rates among the elderly have declined.3 Since the disabled spend more than the nondisabled, some had expected these trends to result in slower growth in Medicare per capita spending. However, this has not occurred; the trends in per capita spending among the disabled and nondisabled have narrowed. In short, per capita spending among the nondisabled Medicare population has risen faster than rates among the disabled population. Our paper provides some insight into why per capita spending among the nondisabled has accelerated.4 Our tabulations from the 1987 National Medical Expenditure Survey (NMES) and the 2002 Medical Expenditure Panel Survey (MEPS) reveal that disability rates among the entire Medicare population declined from 27 percent in 1987 to 22 percent in 2002. At the same time, however, the prevalence of obesity among the nondisabled increased sharply, rising from 12.6 percent in 1987 to 23 percent by 2002. The rising share of obesity among the nondisabled has been associated with a rise in the clinical incidence of several chronic conditions—in particular, type 2 diabetes—as well as rising rates of treatment, as presented in our paper. These trends along with increased longevity have contributed to the rise in the prevalence of treated disease we report in our paper. In short, changes in disability rates and spending are very different from changes in disease prevalence and treatment rates among the nondisabled and health care spending. The two will not coincide if treatment rates are bringing patients with mildly symptomatic or even asymptomatic conditions under medical treatment. Another factor accounting for the rise in the prevalence of treated disease is the ongoing change in clinical definitions of disease. One of these key changes in the clinical definition of disease—metabolic syndrome—constitutes a major portion of our analysis. Although the clinical significance of metabolic syndrome remains controversial, it has increased the number of potential "patients" eligible for treatment over time. Moreover, the definition of what the syndrome comprises continues to evolve. For instance, the original definition developed by the Third Report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (ATP-III) defined the syndrome as the presence of three of the following five conditions: abnormal (impaired) fasting glucose levels, low high-density lipoprotein (HDL) cholesterol, elevated blood pressure, high triglyceride levels, and abdominal obesity.5 The original definition of impaired fasting glucose in the 2001 recommendations was established at 110 mg/dl. However, the American Diabetes Association lowered the definition to 100 mg/dl in 2004.6 This lower impaired fasting glucose level increased the prevalence of metabolic syndrome among adults age twenty and older from 26.5 percent (using the 110 mg/dl standard) to 31.9 percent—a full five-percentage-point increase.7 IN SUM, WE BELIEVE that our line of research in this area has highlighted the critical importance of three key factors underappreciated in previous work on rising health care spending. These three trends—the increase in the incidence of disease (holding constant the clinical definitions), rising rates of medical interventions, and changing clinical definitions of disease—are important factors accounting for the rise in spending. This framework highlights the critical role for primary prevention to reduce the rising incidence of disease. It also establishes an important priority for future research examining the health benefits purchased through higher spending linked to changing clinical definitions of disease and rates of treatment.
Ken Thorpe (kthorpe{at}sph.emory.edu) is the Robert W. Woodruff Professor and chair of the Department of Health Policy and Management, Rollins School of Public Health, Emory University, in Atlanta, Georgia. David Howard is an assistant professor in that department.
| |||||||||||||||||||||||
Top
NOTES
