|
FROM THE FIELDRising Medicare Costs: Are We In Crisis?
Kenneth Thorpe and David Howard provide an important snapshot of the long-term trends in Medicare beneficiaries’ health care spending. They use clever methods to shed new light on increases in per capita spending, convincingly demonstrating that treatment patterns are a major factor. But several puzzles remain. Trends in per capita spending differ from other findings in that Thorpe and Howard find costs rising fastest for those with multiple conditions, and it is unclear what would happen if one took lifetime spending into account. Reliable predictions about long-term trends will require information on any health benefits associated with increases in diagnosis and treatment.
CONCERN OVER FINANCIAL solvency has focused attention on the spending of Medicare beneficiaries. With the baby boomers set to hit retirement age in just a few years, the paper by Kenneth Thorpe and David Howard comes at a critical juncture.1 By cleverly splicing together data from various sources from 1987 and 2002, they have provided an important snapshot of the long-term trends in health care spending. The findings paint a grim picture. Over this fifteen-year period, Medicare beneficiaries’ real total spending doubled. About half of this increase can be attributed directly to population aging as the Medicare population grew. The other half reflects an increase in per capita spending, which—based on the Medical Expenditure Panel Survey (MEPS)—rose about 50 percent, from $5,243 to $7,704, excluding spending by those in institutions.2 One of this paper’s key contributions is to demonstrate the role played by the prevalence of diagnosed disease. In 1987, patients with five or more conditions accounted for 31 percent of the beneficiary population; by 2002, they were 50 percent of the population. Why there is so much more disease is beyond the scope of their paper, but the explanations they highlight—obesity and medical technology—certainly play key roles. The aging of the population also could lead to a worsening constellation of disease that might explain part of this increase in prevalence. But the authors’ most provocative result comes from the innovative way in which they use data from the National Health and Nutrition Examination Survey (NHANES). Using biological measures, they convincingly demonstrate that doctors are now more aggressively treating patients with metabolic syndrome than a decade ago. Clearly treatment patterns are a large part of the story. But several puzzles remain. In particular, the trends in per capita spending are very different for patients with multiple diseases. A closer examination of their Exhibit 3 data demonstrates that real per capita spending for those with five or more chronic conditions rose 32 percent over this period, whereas there was no change for those with fewer than five conditions. This result appears to contrast with the findings of Michael Chernew and colleagues, that per capita spending rose fastest for those with no limitations on activities of daily living.3 The most likely explanation is that spending trends for those age sixty-five and older differ from the Medicare population as a whole, which includes those who qualify because of disabilities. In fact, Thorpe and Howard’s findings from NHANES, which show marked increases in the treatment of metabolic syndrome, also suggest that costs should be rising fastest among those with less-severe disease as new, previously untreated populations seek care. The trends identified by Thorpe and Howard could be even more dramatic once one takes lifetime spending into account. They demonstrate that much of the spending increase can be attributed to more—and more intensive—treatment of cardiovascular disease. If this treatment has resulted in increases in life expectancy, these calculations would understate the importance of heart disease for future Medicare spending. However, if we are spending more because these people are in worse health, then the research of James Lubitz and colleagues is relevant. They find that cumulative Medicare spending is largely invariant to beneficiaries’ health status when they enter Medicare because healthier people live longer.4 So an increase in disease prevalence does not necessarily lead to an increase in costs over a lifetime because of reduced mortality. To predict whether these cost trends will continue, we need to know the underlying causes of rising costs. Research suggests that new medical technologies have played an important role in explaining rising costs.5 And our own research shows that promising technologies will continue to put pressure on spending by the elderly.6 However, a large portion of the increase stems from the attendant increase in life expectancy, which means that people have longer lifetimes in which to accumulate spending. Ultimately, we will need to measure the health benefits (or decrements) associated with the increased diagnosis and treatment documented by Thorpe and Howard. If relatively healthy patients are receiving more diagnoses with no subsequent improvement in health, then these cost increases warrant greater containment efforts. If, however, bringing people with metabolic syndrome into care can forestall serious complications and reduce hospital spending, then we might be undertreating these populations. Ultimately, it will be difficult to distinguish undertreatment from overtreatment, especially since both could lead to increases in costs. Sorting all of this out will likely keep the research community busy for quite awhile.
Dana Goldman (dgoldman{at}rand.org) is corporate chair and director of the Bing Center for Health Economics at RAND and the National Bureau of Economic Research in Santa Monica, California. Neeraj Sood is an associate economist there.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||
Top
NOTES
