Perspective The NEW ENGLA ND JOURNAL of MEDICINEfebruary 14, 2008n engl j med 358;7 www.nejm.org february 14, 2008661such plans involves greater promo-tion of preventive health measures. The first element in Hillary Clin-ton’s plan is to “focus on preven-tion: wellness not sickness.” John Edwards has stated that “study af-ter study shows that primary and preventive care greatly reduces fu-ture health care costs, as well as increasing patients’ health.” Mike Huckabee has said that a focus on prevention “would save countless lives, pain and suffering by the victims of chronic conditions, and billions of dollars.” Barack Obama has argued that “too little is spent on prevention and public health.”Indeed, some evidence does suggest that there are opportuni-ties to save money and improve health through prevention. Prevent-able causes of death, such as to-bacco smoking, poor diet and physical inactivity, and misuse of alcohol have been estimated to be responsible for 900,000 deaths an-nually — nearly 40% of total yearly mortality in the United States.1 Moreover, some of the measures identified by the U.S. Preventive Services Task Force, such as coun-seling adults to quit smoking, screening for colorectal cancer, and providing influenza vacci-nation, reduce mortality either at low cost or at a cost savings.2Sweeping statements about the cost-saving potential of prevention, however, are overreaching. Studies have concluded that preventing ill-ness can in some cases save money but in other cases can add to health care costs.3 For example, screening costs will exceed the savings from avoided treatment in cases in which only a very small fraction of the population would have become ill in the absence of preventive measures. Preventive measures that do not save money may or may not represent cost-effective care (i.e., good value for the resources ex-pended). Whether any preventive measure saves money or is a rea-sonable investment despite adding to costs depends entirely on the particular intervention and the spe-cific population in question. For example, drugs used to treat high cholesterol yield much greater value for the money if the targeted popu-lation is at high risk for coronary heart disease, and the efficiency of cancer screening can depend heav-ily on both the frequency of the screening and the level of cancer risk in the screened population.4The focus on prevention as a key source of cost savings in health care also sidesteps the question of whether such measures are gener-ally more promising and efficient Does Preventive Care Save Money? Health Economics and the Presidential CandidatesJoshua T. Cohen, Ph.D., Peter J. Neumann, Sc.D., and Milton C. Weinstein, Ph.D.With health care once again a leading issue in a presidential race, candidates have offered plans for controlling spiraling costs while enhanc-ing the quality of care. A popular component of Copyright © 2008 Massachusetts Medical Society. All rights reserved. Downloaded from www.nejm.org at UNIVERSITY OF MINNESOTA on February 19, 2008 .PE RSP ECT IVEn engl j med 358;7 www.nejm.org february 14, 2008662than the treatment of existing con-ditions. Researchers have found that although high-technology treatments for existing conditions can be expensive, such measures may, in certain circumstances, also represent an efficient use of re-sources.5 It is important to analyze the costs and benefits of specific interventions.A systematic review of the cost-effectiveness literature sheds light on these issues. We analyzed the contents of the Tufts–New England Medical Center Cost-Effectiveness Analysis Registry (www.tufts-nemc.org/cearegistry), which consists of detailed abstracted information on published cost-effectiveness studies through 2005. Each registry article estimates the cost-effectiveness of one or more interventions as the incremental costs (converted here to 2006 U.S. dollars) divided by the incremental health benefits quanti-fied in terms of quality-adjusted life-years (QALYs). Low cost-effec-tiveness ratios are “favorable” because they indicate that incre-mental QALYs can be accrued inexpensively. An intervention is “cost-saving” if it reduces costs while improving health. Poorly per-forming interventions can both in-crease costs and worsen health.Our analysis was restricted to the 599 articles (and 1500 ratios) published between 2000 and 2005 that properly discounted future costs and benefits. We classified 279 ratios as preventive because they refer to interventions designed to avert disease or injury; all 1221 other ratios pertain to treatments, a category that includes both “ter-tiary” measures (designed to ame-liorate the effects of a disease or condition) and “secondary preven-tion” measures (designed to re-verse or retard progression of an existing condition), such as the use of implantable cardioverter–defi-brillators in patients with myo-cardial disease.The bar graph shows that the distributions of cost-effectiveness ratios for preventive measures and treatments are very similar — in other words, opportunities for ef-ficient investment in health care programs are roughly equal for prevention and treatment, at least as reflected in the literature we re-viewed. Moreover, both distribu-tions span the full range of cost-effectiveness. The table shows the cost-effectiveness ratios for select-ed interventions of various types.These results are consistent with earlier reviews but cover a larger sample of studies and quantify benefits in terms of QALYs. Some preventive measures save money, while others do not, although they may still be worthwhile because they confer substantial health ben-efits relative to their cost. In con-trast, some preventive measures are expensive given the health ben-efits they confer. In general, wheth-er a particular preventive measure represents good value or poor value depends on factors such as the population targeted, with measures targeting higher-risk populations typically being the most efficient. In the case of screening, efficiency also depends on frequency (more frequent screening confers greater benefits but is less efficient). Third, as is the case for preventive measures, treatments can be rel-atively efficient or inefficient.Of course, our review reflects a selected sample of