Economic Evaluation Across the Four Faces of Prevention: A Canadian Perspective

In 1986, Louise Russell published her landmark book, “Is Prevention Better Than Cure?”, in which she evaluated the health and economic benefits of preventive health care interventions and tested the validity of the common assumption that prevention saves money. While debunking the myth that prevention is invariably cost-saving, Russell insisted that prevention is only rightfully judged on whether it is a worthwhile investment in health, rather than on its cost-saving potential. Almost three decades later, the notion that “an ounce of prevention is worth a pound of cure” still grips the imagination of policymakers and members of the public. We were commissioned by the Canadian Medical Association to review the economic evaluation evidence on prevention in the hope that such a review would assist health and health care priority setting in Canada. Prevention Versus Cure? In discussions of health policy and resource allocation, prevention is often pitted against cure and illness care. Prevention and illness care are not inherently competitive for resources. They serve different objectives and respond to different needs. In the quest for resources, prevention faces a difficult challenge in obtaining public and political support. In contrast to illness care, prevention has no identifiable beneficiaries and is usually characterized by immediate costs and delayed benefits. Economic Evaluation and its Use in Priority Setting It is widely argued that evidence of efficiency derived from economic evaluations should play an important role in health care priority setting and coverage decisions. Indeed, to our knowledge, every recently proposed priority-setting framework includes efficiency assessment as a crucial step. A limited number of jurisdictions require and use evidence of efficiency as part of coverage decision-making within public health care insurance programs. These examples are notable, however, precisely because they are exceptions. Overall, the use of economic evaluation evidence in priority setting lags far behind both the prescriptions of priority-setting frameworks and the expectations of many health policy makers and analysts. A number of factors are likely responsible for this. Some stem from the fact that integration of economic evaluation evidence into decision-making is neither simple nor straightforward. But more fundamentally, most health care interventions have never been subject to an economic evaluation and the interventions that have been assessed tend to be those that are most easily studied (rather than those for which the need for such evidence is most pressing). Many more economic evaluations exist for clinical prevention interventions, for example, with more easily defined populations, interventions, and settings, and more easily measured outcomes, than for interventions drawn from the areas of health promotion, health protection or healthy public policies. To the extent that priority-setting exercises restrict consideration to interventions for which such evidence exists, only a limited and non-representative set of interventions can be considered. One hope, of course, is that in the long run demand by those setting priorities and making coverage decisions will bring forth a larger supply of such studies. Economic Evaluation Evidence for Preventive Interventions. Deciding which topics to include in this review was a formidable task. While economic evaluation of preventive activities is not as frequent as for treatment, the volume of the prevention literature is vast. (Initial explorations of potentially relevant literature for this paper, for example, yielded over 5000 articles.) To guide our work, we identified 290 recommended prevention interventions and an additional 23 preventive interventions with potentially large population health impacts. We found no economic evaluations for 159 of the 290 recommended interventions (55 percent). Our literature identified 672 economic evaluations of the remaining 154 preventive interventions. The majority (55 percent) evaluated clinical prevention interventions. The next largest group of evaluations assessed health protection interventions (31 percent), a significant proportion of which were universal or mandatory screening or immunization programs. Health promotion interventions represented 12 percent of the evaluations and healthy public policy interventions represented 2 percent. The volume of relevant economic evaluations was far greater than we anticipated at the outset of the project. Accordingly, after categorizing available economic evaluations of recommended preventive interventions by type of intervention, the condition or issue targeted by the intervention, the study setting, and the availability and strength of effectiveness evidence, we summarized the results of economic evaluations of a sample of five diverse interventions that are not universally provided in Canada and for which a considerable body of economic evaluation evidence is available. Synthesis of Economic Evaluation Evidence for Five Selected Preventive Interventions • Varicella vaccination • Colorectal cancer screening using fecal occult blood testing (FOBT) • Needle exchange programs • Community water fluoridation • Day care or preschool programs In summarizing and interpreting the results of economic evaluation evidence for these five syntheses, we addressed three questions: • Does the intervention produce a net benefit from the societal perspective? • Is the intervention cost-saving from the payer perspective? • Where cost-benefit from a societal perspective has not been adequately assessed and the intervention is not cost-saving from the payer perspective, might the intervention nonetheless be a worthwhile investment in health (i.e., give value for money)? For each intervention we found a high degree of consistency among economic evaluation studies, despite differences in methods and settings. In particular, we did not observe systematic differences in findings between economic evaluations set inside or outside of Canada. All of the interventions we examined produce a net benefit to society. Needle exchange programs and water fluoridation are also cost-saving from a payer perspective. In both cases, there are sometimes multiple payers, which means that program costs may be born primarily by one payer while another payer may be the principal beneficiary of cost-savings resulting from the intervention (e.g., reduced treatment costs). The remaining interventions—varicella vaccination, colorectal cancer screening with FOBT, and day care or preschool programs—while not cost saving from the payer perspective (with the possible exception of preteen varicella vaccination), may still be sound investments in health. Decisions about whether to make those investments will appropriately depend on a variety of factors, some related to and others external to the economic evaluation evidence. Next Steps. A large volume of unappraised and unsynthesized economic evaluation evidence is available for many preventive interventions. On the other hand, economic evaluation evidence is completely lacking for the majority of recommended preventive interventions. If economic evaluation evidence on prevention is to be used to assist health and health care priority setting in Canada, the gaps that we have identified need to be filled. Critical activities include: • Systematic reviews of effectiveness evidence for health promotion, health protection, and healthy public policy interventions • Economic evaluations of individual preventive interventions for which economic evaluation evidence is currently lacking • Systematic reviews of economic evaluation evidence for effective preventive interventions These activities will require substantial resources. Significant work in this area has been and continues to be undertaken outside of Canada, particularly in the United States. Application of Economic Evaluation Evidence to Policy Decision Making. Policy decision making that incorporates economic evaluation evidence cannot be reduced to rank ordering of programs by summary measures of efficiency and the mechanical application of thresholds to determine which programs will be implemented or continued. Even if such an approach were desirable, its feasibility is questionable given that economic evaluation evidence based on a common metric and common comparator is unlikely to be available across a full range of programs under consideration. Decisions regarding public investments in health care programs are inevitably influenced by a variety of factors—some economic, some political, and some having to do with social values and preferences. These include: • Differential timing of costs and benefits • Opportunity costs • Availability of required technology and human resources • Program scope • Nature of benefits • Who benefits • Who pays Economic evaluation evidence can make a useful contribution to policy decision making. Given the dearth of economic evaluation evidence for preventive interventions, however, it is not reasonable to make such formal evidence a pre-requisite for policy action. Requiring economic evidence as a mandatory input to decision making would, in the short term, delay the implementation of preventive programs with demonstrated large population health effects that had not yet been subjected to economic evaluation. Perhaps more importantly, in the long term such a requirement would discriminate against health promotion, health protection and healthy public policy interventions whose costs and consequences are often difficult to measure credibly because they are spread across multiple health and social domains. In spite of these challenges, we hope that this review demonstrates the value of increasing the use of economic evaluation methods to inform decision making for preventive interventions.