How Much Will We Pay for a Year of Life?
STANFORD GRADUATE SCHOOL OF BUSINESS—The single biggest factor contributing to the astronomically rising cost of healthcare is the emergence of expensive new technology. Science is finding more effective ways of treating diseases and extending life, but at a substantial cost. How much are we—and should we be—willing to pay, as a society, for improving health outcomes?
New research from Stanford and Wharton shows that $50,000—the average figure used internationally as a “threshold” for making medical allocation decisions—is low. A more realistic figure is probably a minimum of $129,000, which represents what it would cost to give a person an additional “quality-of-life adjusted” year of life. Moreover, the researchers argue, making decisions on whether insurance should cover medical interventions based on their cost-effectiveness leads to profound ethical dilemmas.
Outside the United States, countries such as the United Kingdom and Australia that offer national health care have developed explicit systems to determine the overall cost-effectiveness of a new medical intervention. This includes calculating the incremental cost of a treatment against the incremental improvements in the patient’s health, and comparing that figure to a threshold number. “As long as the ratio is below that threshold number, a given treatment is accepted as part of the healthcare offerings; otherwise, it’s rejected,” explains Stefanos Zenios, the Charles A. Holloway Professor of Operations, Information, and Technology at the Stanford Graduate School of Business and one of the coauthors of a new paper forthcoming in the journal Value in Health. The paper is coauthored with Glenn Chertow, professor of nephrology at the Stanford School of Medicine and Chris Lee of the Wharton Business School.
The U.S. Medicare system—the national health insurance system for the elderly—has thus far eschewed making allocation decisions based on such bald numerical calculations. “Big decisions typically are made on the basis that the treatment is ‘medically necessary and appropriate,’ but that concept is vague,” says Zenios—and it does not include explicit cost considerations. Debate on the cost of the new Medicare prescription drug benefit program suggests that making coverage based on clinical evidence alone without attention to cost may not be feasible in the long term.
The authors wanted to know: If Medicare were to begin accepting or rejecting coverage on treatments made available by new technologies based on cost, what might a realistic threshold look like? To make such a determination, they ran computer models using data from more than half a million patients who underwent kidney dialysis—an expensive procedure covered under Medicare that has typically been used as a benchmark for evaluating the cost-effectiveness of all new technologies internationally. “We found that starting dialysis earlier than the current practice allowed by Medicare would be more expensive, but would likely be associated with longer life and fewer medical complications,” Zenios said.
In such a case, the average incremental cost was approximately $129,000 for a “quality-adjusted” year of life. “This means that if Medicare paid an additional $129,000 to treat a group of patients, on average, that group would get a total of one more quality-adjusted life year,” Zenios said. Based on surveys of patients with kidney failure, one “quality of life” year is deemed the equivalent of about two years of life under dialysis.
Medicare could consider using the $129,000 figure as a benchmark to determine whether to cover treatments using new technologies. “Say a new type of treatment for cancer comes along,” explains Zenios. “If the incremental cost of that new technology was more than $129,000 for a quality-of-life adjusted year, then the recommendation would be that Medicare not cover that new technology.”
But the research comes with warnings. “The first caveat is that the dialysis threshold may not be an appropriate benchmark for all technologies,” says Zenios. “This is something that should be debated,” said Zenios.
The second caveat, Zenios says, is that Medicare would quickly be faced with a host of ethical concerns if it started applying the $129,000 threshold differentially to selective groups. For the sickest patients, the researchers determined that the average cost of an additional quality-of-life year was much higher: $488,000. “Applying the $129,000 threshold in a very sharp way for specific groups and individuals would mean that the sickest subgroups of patients would be denied access to expensive treatments such as dialysis.”
For some policy makers, such a decision might be tempting, given that the sickest benefit the least from dialysis, and if the threshold were raised to even half that, or $240,000 per quality-of-life year, 90 percent of patients could be treated.
"Dialysis is a life-sustaining therapy,” co-author Glenn Chertow says. “While kidney transplantation is usually preferable, waiting times for deceased donor organs are often longer than six years. Without dialysis, hundreds of thousands of persons in the U.S. would die each year due to complications of kidney failure. However, it is difficult to justify the burden and expense of dialysis when persons have other serious health conditions, for example, advanced dementia or cancer. In these settings, dialysis is unlikely to provide any meaningful benefit."
The issue of treating patients differently depending on their health poses ethical questions. “We argue that any thresholds should be applied in keeping with the principles of social justice established by the American political philosopher John Rawls, who held that resources should be allocated to benefit everyone, including the most vulnerable individuals,” says Zenios. According to this philosophy, the authors maintain that the average figure—$129,000—could be used as a global figure for the purposes of accepting or rejecting an entire technology for coverage, but not for accepting or rejecting specific subgroups once that technology has been approved.
How much money we should allocate to helping the sick, sicker, and sickest will only continue to be debated ever more vociferously as newer, costlier technologies continue to emerge. “We need to be conscious about the social and ethical implications of any numerical figures arrived at by this or any other study,” concludes Zenios.