“Value for Money”: Use With Care FREE

In this issue, Owens and colleagues (1) claim that the goal of policy should be to “preserve the delivery of interventions that do provide good value.” They also acknowledge that evaluating costs without considering value is problematic because it could lead to “inadvertent reductions in effective and efficient care.” Cutler and colleagues (2) have argued that an intervention that yields more benefits than costs is socially beneficial regardless of whether it increases spending. The American College of Physicians Clinical Guidelines Committee makes an important contribution to this policy debate by offering definitions and concepts to guide the search for high-value health care, but the “value-for-money” goal raises many questions that warrant further discussion.

Owens and colleagues acknowledge that evidence about the effectiveness of health interventions is lacking, and that our ability to assess the quality of life associated with particular health states is inadequate. The Obama administration's investment in comparative effectiveness research provides some hope that we will improve the evidence base, but the checkered history of such research in the United States should give us pause (3). The private sector may use new evidence about effectiveness to control costs, but because the health reform law explicitly forbids the Patient-Centered Outcomes Research Institute from considering costs, it will be difficult for the agency to produce the evidence these recommendations require.

Even if we overcome the absence of relevant data, it may be more difficult to overcome the absence of consensus about which costs and benefits to count in the first place. Owens and colleagues argue that it is important to account for downstream costs and benefits when assessing health interventions. But which downstream costs and benefits should stakeholders include in their analysis? Should balance sheets account for the costs and benefits of “unrelated illness and non–health care consumption and production during added life years,” as suggested by Meltzer and Johanesson (4)? What sort of discount rate should we use? If we discount the future too heavily, we may place unfair burdens on future generations. These difficult issues hide within the apparently simple concept of value for money.

What may be a good value for money for a sick person may not be good value for other members of society; for example, a dying patient and his or her family may judge the high cost of a new intervention to be worth the expense. Formal economic evaluations try to address this issue by adopting explicit standards that specify and place limits on the economic value of health benefits. Yet, as Owens and colleagues acknowledge, setting the standard is crucial and problematic. In the United States, policymakers have explicitly rejected attempts to set these limits by using economic evaluations (5). Suggesting that the economic value of a new cancer drug that adds a few months to life may not be worth the cost is politically unacceptable. Even if we had the political will to incorporate explicit standards for the economic value of health benefits into the evaluation of medical technology, economists do not agree on which standard to use (6 - 7).

In the United Kingdom, the National Institute for Health and Clinical Excellence (NICE) uses a cost-per-quality-adjusted life-year (QALY) methodology for evaluating health interventions. NICE does not have an explicit policy regarding the threshold for cost-effectiveness, but its studies tend to use a cost-per-QALY threshold between £20 000 (about $30 000) and £30 000 (about $42 000), which is lower than that suggested by Owens and colleagues. In addition, Owens and colleagues imply that the NICE recommendations have reduced costs; however, NICE has approved most of the technologies it has reviewed (8). Using a higher cost-per-QALY threshold would eliminate even fewer interventions.

The application of these estimates in economic evaluations may be valuable if it forces decision makers to be explicit about what are often implicit and unexamined choices concealed within policy decisions (9 - 10), but the variations in estimates raise serious concerns. Using labor market data and “willingness-to-pay” surveys to better understand the value patients place on health gains through so-called “revealed preferences” (11) is laudable, but these studies do not clarify matters. Instead, they risk presenting value claims as the result of a technical, objective exercise rather than a claim about values.

Perhaps the biggest problem with cost–utility analysis in health care is that expenditures on health care cannot be compared with other societal needs. In the context of the United Kingdom's National Health Service, the failure to consider opportunity cost may eliminate existing, but unassessed, health care technologies and services that are a better value than the “cost-effective” technology included in these assessments (12). In the United States, which does not use a budget to control health care spending, the value-for-money standard alone is insufficient to limit spending.

Owens and colleagues recognize that efforts to calculate value for money do not tell us how much to reduce health care spending. Because they presume that cost control is inevitable, they focus exclusively on how to make those cuts while maintaining high-quality care. Unfortunately, this ignores the opportunity costs that flow from applying the value-for-money principle in a country that has never mustered the political will to say “no” to additional health care spending. Preserving “high-value, high-quality care” while slowing the rate of cost increases will be a highwire act in the face of a steady stream of new and improved technologies, most of which may increase costs.