WITH SUBSTANTIAL SUPPORT ACROSS THE POLITIcal spectrum, the Obama administration has included in the American Recovery and Reinvestment Act more than $1 billion to support comparative effectiveness research. At the same time, the president has demanded reforms in the US health care system to make health care more affordable for all US citizens. This Commentary focuses on the interaction of these 2 initiatives: what will be the cost effect of spending $1 billion on comparative effectiveness research? First, it is important to understand what comparative effectiveness research will include. Discussions to date suggest that most of the funds will be spent comparing one clinical procedure, device, or drug with another. The funds are less likely to be spent testing the comparative effectiveness of one way of paying for care vs another, of organizing care using a chronic disease model vs another organizational principle, or of implementing aggressive disease management programs. Countless studies could be conducted to assess the comparative effectiveness of clinical procedures, devices, or drugs. Consider, for instance, a company that manufactures an improved surgical needle. Comparing that needle to the needle currently used could be included in comparative effectiveness research. A nearly infinite number of studies could be conducted to determine how often a person with back pain should receive chiropractic treatment, how often a patient with hypertension should receive follow-up care, how often a patient should obtain a dental checkup, or what form of radiation and chemotherapy will achieve the best outcome for a patient with cancer. Since the government is using taxpayer dollars to fund the comparative effectiveness initiative, it would be appropriate to have an organizing principle to guide the selection of which aspects of medical care to examine. Without some overarching principle, the program is likely to become a free-for-all or a full employment program for health services researchers and epidemiologists. Researchers would submit proposals to study drug A vs B, or procedure A vs drug B. A study section would assess the proposals and decide which ones to fund, using the standard framework: the comparison should be important, the design should answer the question, and the study should be feasible to conduct. But such an approach pays no attention to achieving the president’s objective of making care more affordable or to insisting that the results of the comparative effectiveness studies be implemented quickly. So it is conceivable that most of the $1 billion could be allocated, for example, to testing a treatment that is more expensive than standard therapy to determine if the new treatment produces slightly better health. There is nothing wrong with such studies; the question is whether the scarce resources allocated in the stimulus package should be used in this way. It is clearly in the interest of companies that develop new treatments (eg, devices, drugs) to compare them with current or standard therapy. Why should taxpayer dollars subsidize that comparison and facilitate rapid entrance into the market of treatments that are slightly better but substantially more costly? The comparative effectiveness funds should be allocated according to a framework designed to identify procedures, devices, and drugs that would reduce cost but not diminish health. This would help achieve the goal of making health care more affordable. Such a framework would have 2 new required features. First, a grant or contract to spend public money must include an initial analysis to establish a business case that implementing whatever is being proposed would reduce the cost of care by a certain percentage. Second, it would not be enough to establish that implementing thenewdrug,device,orother therapywill savemoney. The history of science shows that it takes a long time for new knowledge to be incorporated into day-to-day practice. So a second requirement for work funded under the stimulus package should be that successful innovations are implemented immediately. Thus, a successful application under the comparative effectiveness initiative must include constituents, such as health care organizations, hospitals, physicians, or organized community groups, that would agree to adopt the new therapy immediately if it were shown to be as safe as the old therapy but substantially less expensive. If such constituent letters were not contained in the grant proposal or contract under which the research would be conducted, the money should be spent investigating other topics. Using comparative effectiveness funds correctly could foster a sea change in the way industry looks at product development. To return to the example of the new needle, the goal would no longer be to conduct a study to convince practitioners to pay more for a better needle. Rather, the goals would be to make better and safer needles, reduce the price of needles, and ensure that the new needles are adopted im-