Cost-effectiveness analysis (CEA) is one tool available to decisionmakers to help determine the relative value of different approaches to improving health or life expectancy. For a variety of reasons, however, CEA has not been widely used in health policy settings. This study represents a review of the methodological and theoretical aspects of CEA, its effectiveness, and its expected outcomes and uses. The resulting report is a consensus among many of the top researchers in the field and provides guidance to the conduct of CEA. The report sets forth several overlapping recommendations: some intended to improve the overall quality of CEA and some intended to set a research agenda for further improvement of CEA. Recommendations are made in the areas of framing and designing the study, identifying and valuing outcomes, assessing effectiveness, estimating costs, discounting, reflecting analytical uncertainty, and reporting analyses. It also discusses practical constraints to CEA's policy uses, carefully taking into account the needs of its varied audiences.
The purpose of this study was to assess the current state of the art of the use of CEA in health care and to recommend steps to improve its quality and comparability. The Panel on Cost-Effectiveness in Health and Medicine, working closely with methodologists and liaisons from the Public Health Service and the Health Care Financing Administration attempted to identify a number of methodological problems in existing CEA's that limit the analyses' policy relevance and usefulness.
Society is confronting many difficult choices regarding the provision of health care services and the funding of public health programs to improve the Nation's well-being. The aging of the population and the costs of expanding medical technologies have placed growing pressure on health care resources. The private sector, which has long paid for health insurance for a significant proportion of the population, is confronted by competitive challenges and shifting nationwide economic trends. In the public sector, there are competing demands for tax dollars to fund health programs that range from the clinical care supported by Medicare and Medicaid to the population-based approaches employed by public health programs. The Nation is faced with an apparent series of tradeoffs among priorities, as we seek to make wise investment decisions to improve health.
There are many ways to resolve these choices, none of them straightforward and none without compromise. To make the most informed decisions, information is needed about the impact of programs and interventions, their costs, and what is given up when one program is implemented at the cost of another. This is true for Government, managed care organizations, providers, employers, and consumers. As the pressures to control U.S. health care spending have accelerated, the term "cost-effective" has found an increasingly broad and interested audience. But this term often has different meanings to groups as disparate as the U.S. Congress, business leaders, managed care organizations, the pharmaceutical industry, and the public. A more systematic and reliable approach to determining relative values through CEA is required.
This report summarizes the work of the Panel on Cost-Effectiveness in Health and Medicine, a non-Federal expert panel appointed in 1993 by PHS. The panel consisted of 13 scientists and scholars with expertise in CEA and related methods. The panel was charged with assessing the current state of the science of CEA and with providing recommendations for the conduct of studies to improve its policy relevance and utility. The panel met from 1993 through 1995 in collaboration with methodologists from Public Health Service agencies and the Health Care Financing Administration.
The panel organized itself into nine working groups focusing on specific topics: (1) the role of CEA in decision making; (2) CEA's theoretical foundations; (3) the appropriate framing of a study; (4) measuring effectiveness; (5) valuing outcomes; (6) measuring costs; (7) discounting considerations; (8) handling uncertainty within a study; and (9) guidelines for reporting analyses. Panel members and staff drafted papers on topics addressing the major controversies in each area. The papers contained recommendations that were debated by panel members in successive meetings until consensus was reached and formal recommendations could be issued or, in a few cases, until it became clear that consensus could not be reached. In areas where consensus was elusive, the report describes the full range of arguments presented.
CEA furnishes a framework for evaluating the economic and health impacts of different types of investments and can inform a wide range of policy decisions. The results of a CEA are typically summarized as a cost-effective ratio, for example, $138,000 per quality-adjusted life-year, in which the denominator reflects the gain in health from a specific intervention (e.g., years of healthy life gained, premature births averted, or sight-years gained) and the numerator reflects the cost in dollars of obtaining that gain.
A well-constructed CEA provides information that helps decisionmakers identify which strategies will best serve their programmatic and financial objectives. Depending on the type of analysis conducted, a CEA can compare similar or very different approaches to a given problem, such as strategies for decreasing cardiovascular disease. For example, State and Federal decisionmakers with responsibility for Medicaid programs, as well as for public health systems, may wish to know where investments of public dollars will have the largest impact in decreasing premature death from heart disease. In the private sector, a managed care organization might wish to know how extending its coverage to include cardiovascular disease prevention programs would affect the organization's expenditures.
Some decisionmakers may also wish to make comparisons among health conditions and programs; for example, they may wish to know the efficiencies of investments in interventions to improve neonatal outcomes relative to investments targeted at decreasing cardiovascular disease in adults. Ideally, CEA should allow its users to assess the relative value of dissimilar health-producing programs or treatments.
Unfortunately many CEA's are inadequate to these tasks. Reviews of CEA's chronicle a troublesome failure of CEA practitioners to adhere to basic analytic principles; there also is wide variation in sources and the quality of information on costs and effectiveness. Concerns about the discretionary nature of the application of analytic methods and the economic incentives of commercial sponsors of some analyses have resulted in a decision by at least one medical journal to restrict publication of any CEA where authors may have a financial conflict of interest.
Methodological inconsistency has often stymied the application of CEA in important public policy decisions. For example, in the Oregon Medicaid experiment, in which the State legislature attempted to expand coverage to uninsured Oregonians by basing inclusion of services on the cost-effectiveness of different treatment-condition pairs, deficiencies in CEA technique and knowledge were partly responsible for the fact that the initial list of included services lacked credibility. In a more narrow application to policy needs, a review by the National Cancer Institute of an extensive literature on the cost-effectiveness of screening mammography found that the study results ranged from a finding that mammography would be cost saving to a finding that it would cost nearly $84,000 per year of life saved. Confusing results can lead to confusing policy.
The imprecision attached to the term "cost-effective" stems in part from the variety of masters the concept serves: purchasers of health care, who use the term to convey a careful assessment of the value of different health care services; producers of health care technologies and programs, who use the idea to support marketing claims; and advocates for particular constituencies or illnesses, who use the term to garner resource investments. But the imprecision also comes from the manner in which methods of performing CEA have evolved over the past three decades. CEA is an analytic tool whose fundamental purpose is to provide information to decisionmakers about the relative value of different approaches to improving health, life expectancy, or both. Architects of the field and analysts who apply CEA methods come from a number of academic disciplines, including economics, medicine, operations research, medical sociology, psychology, public health, and ethics. Each discipline brings a particular set of concepts and a unique language that have been melded in the building of the CEA technique.
Use of Results
The panel has produced a book, Cost-Effectiveness in Health and Medicine, summarizing CEA methodology with the objective of improving its usefulness to policymakers. The report will be distributed to Federal decisionmakers and analysts to inform them about the interpretation of existing CEA and ways to improve studies conducted by and for the Government. The interdepartmental group is planning broader dissemination and discussion of the panel's work through an international conference. The conference will explore the implications of the panel's recommendations and will identify next steps in moving toward the production and use of high-quality comparable CEA.
Office of Disease Prevention and Health Promotion
PIC ID: 5619
Office of Disease Prevention and Health Promotion, Washington, DC