Essentials of Economic Evaluation in Health Care - Term Paper Example

Introduction Health care programs require funding for its implementation and continued operation. For funding to be possible, these programs must undergo evaluation regarding its effectiveness and efficiency. The question, however, is how to measure these criteria. Questions of optimal health care programs and procedures are vexing not only for clinicians but also for those who organize and pay for health care. Formulating rational policies in clinical medicine is an extraordinarily complex and difficult undertaking. Technology evolves quickly, effectiveness data are often incomplete and the ethical implications of our decisions are not straightforward. In addition, the cost of health care is assuming a larger role in clinical policymaking. Even wealthy nations, squeezed between the Scylla of exponential technology growth and the Charybdis of shrinking health care resources must make hard choices about which technologies they can afford (Gold et al, 2007). Economic evaluation is a methodology that is being used increasingly to help clinicians and policymakers think about these hard choices. Economic evaluation is the “comparative analysis of alternative courses of action in terms of both their costs and consequences” (Drummond et al, 2005). When measuring the effects of health programs on both resource consumption and health, economic evaluation allows us to see how much health care “bang” we are getting for our “buck.” Over the last 20 years the number of published CEAs has grown exponentially, particularly in clinical journals. Insurers, health maintenance organizations (HMOs), and regional and national governments are showing an increasing interest in the results of CEAs, particularly in the area of pharmaceuticals. Economic evaluation has matured as a scientific discipline. In this brief review we discuss some of the characteristics and uses of economic evaluation and provide direction for those who wish to learn more. Comparisons between Programs Full economic evaluations are always comparative. No treatment is ever “cost-effective” in isolation. It is always more or less so in relation to some other treatment, even if the alternate treatment is to “do nothing.” Even doing nothing has health and cost consequences that must be considered (Ketley, 2009). What programs should be compared? In their work on the subject, Zweifel and his colleagues (2009) states that economic evaluation ideally includes all potential and mutually exclusive programs for a defined population. Thus an ideal economic evaluation of colorectal cancer screening, for example, involves a comparison of all methods of stool occult blood testing, sigmoidoscopy, colonoscopy, and barium enema—singly and in every combination over all possible screening intervals which is clearly a Herculean task. Fortunately, study design and modelling limitations (tractability) limit the consideration of therapeutic alternatives. Economic evaluations carried out in parallel with randomized controlled trials (RCTs), for example, may be limited to the examination of only two or three alternatives. There is no widely accepted formal approach to determining which subset of possible alternatives should be examined. One set of guidelines recommends that the least expensive, most effective, and most widely used strategies be evaluated (Ketley, 2009). Methods of Economic Evaluation There are essentially four methods used in economic evaluation of health care. They are (Drummond et al, 2005): Cost Minimization Study Cost Effectiveness Analysis Cost Utility Analysis Cost Benefit Analysis The way in which health effects are measured is the principal factor in determining the design of the economic analysis. The simplest way to deal with health effects is to ignore them. This is perfectly legitimate in the rare situation where two programs are of equal efficacy (cost minimization study). Similarly, a simple tabulation of costs (cost comparison) may show that one program is preferred if it is less costly and is known to produce greater health benefit (dominant strategy). Unfortunately, partial evaluations (those that include only costs) are rarely sufficient. It is usually necessary to measure health effects as well as costs for a full evaluation. Health effects may be measured in natural clinical units (e.g., aneurysm ruptures prevented, cancers detected, lives saved), in which case the analysis is a cost-effectiveness analysis. Gold et al (2007) argues that the limitation of this approach is that the denominator is expressed using different measurement units in each study, so comparison across different disease states is difficult. An alternative approach is to convert health outputs into standardized units such as quality-adjusted life years (QALYs) or healthy-year equivalents (HYE). This type of analysis is known as cost-utility analysis. The chief virtue of this approach is that it allows comparisons among programs with dissimilar health outputs, both within and across disease states. Thus it is feasible to compare the economic attractiveness of screening mammography with breast cancer treatment or in-home haemodialysis and bone marrow transplantation for acute leukaemia by determining the incremental cost of producing a standardized health unit in each program (Zweifel, 2009). The most widely used unit of health output is the quality adjusted life-year (QALY). In this conceptualization, health has two dimensions: length of life and quality of life. Utility, a measure of patient preference, is used as a weight to adjust length of life for varying quality and produce a standardized unit of health (Shepard and Weinstein, 2005). The final approach to characterizing health benefits is to express their value in monetary terms (cost-benefit analysis). This approach dramatically simplifies interpretation of the analysis, as costs and health effects are expressed in the same units (i.e., dollars). A net cost saving or net cost increase suggests that a program is economically attractive or unattractive. This approach has always held more attraction for economists than health care decision makers because it requires the analyst to value health and life in monetary terms, a step that often engenders distrust among consumers of the research. Nonetheless, cost-effectiveness and cost-utility analysis also require implicit valuation of health outcomes in monetary terms if they are used to direct policy decisions (Elliot and Payne, 2005). Thus under most circumstances, all three forms of full economic evaluation are equivalent (though not equally informative), despite differences in reporting style. Cost Categories The cost of any health care intervention is the value of the resources required to deliver or provide it. This seems straightforward, but measuring costs is difficult. Charges associated with health services are much easier to obtain than costs and are frequently used in economic analyses. Ideally, economic evaluations should utilize actual costs drawn from a large and representative sample. In practice, there is often a trade-off between accuracy of costing and the representativeness of the sample from which the costs are derived. Briggs (2006) and Detsky and Nagle (2006) offers the following type of cost used in economic evaluation studies. Direct costs are expenditures directly induced or averted by an intervention. Direct costs are related to provision of a medical service (direct health care costs). Examples in surgery are the costs associated with imaging studies, surgery, drugs, and follow- up care, as well as the cost of care prevented or induced as the result of early or late treatment. Patients must travel to their appointments and may incur lodging expenses; or they may have expenses related to special diets, clothing, or prostheses. These costs are direct in that they are expenditures induced by the intervention, but they are nonmedical and hence are termed direct nonmedical or direct non-healthcare costs. Illness and treatment result in productivity loss as well as direct costs. Individuals who are ill or who die prematurely cannot work and contribute productively to the economy. The concept of indirect cost, then, refers to the productivity loss caused by disease related morbidity and premature mortality. Productivity losses caused by disease represent real costs but there are no universally agreed on method for valuing these losses. Some authors have argued that the impact of productivity loss is already accounted for in measures of health benefit (the denominator of the cost-effectiveness ratio) and therefore should not be assigned a dollar value. The practical difficulty of measuring productivity loss and the methodological disagreement about whether and exactly how to do it accounts for the fact that few economic analyses include indirect costs. Intangible cost is the value of pain, suffering, grief, and other nonfinancial outcomes associated with illness. Though it is possible to express the “cost” of pain and suffering by attaching a monetary value (as is done for a cost-benefit analysis), it is more common to express intangible costs as health outcomes. Thus intangible costs are usually represented in the denominator of cost-effectiveness or cost-utility ratios. Perspective of the Analysis Which costs should be included in the analysis is determined by the perspective of the analysis. Economic evaluations may adopt the perspective of society, government, government agency, or other third-party payer, hospital, department, individual health care provider, or patient. The societal perspective is the broadest perspective and takes account of all costs, irrespective of their distribution. Analyses that adopt a societal perspective should include, in addition to direct medical costs, all productivity losses, direct nonmedical costs, and costs to individuals. The most frequently utilized perspective is that of the payer (government or other third party), which excludes costs borne by individuals for direct health care, insurance premiums, direct nonhealth costs, and indirect costs (Rice et al, 2005). This is arguably the most relevant perspective for decision making, as decision makers within health care organizations attempt to maximize health outputs subject to their own particular budgetary constraint. The most comprehensive approach is to carry out the analysis from all perspectives, including societal and individual perspectives. If the optimal decision differs according to the perspective of the analysis, it is brought into sharp relief by the analysis and can be taken explicitly into account by the decision maker/s (Briggs et al, 2005). Evaluating Uncertainty: Sensitivity Analysis, New Statistical Approaches Sensitivity analysis is the main tool analysts use to evaluate whether the qualitative conclusion reached in economic analyses, particularly those based on decision models, are robust to the uncertainties in the model. Sensitivity analysis involves varying one or several parameters across the range of uncertainty to determine whether the analytic result changes. “Structural” sensitivity analysis involves changes in model structure to determine if modelling assumptions are creating bias. Analysis of extremes is another frequently employed technique: By varying multiple parameters simultaneously, the model is systematically biased toward and then against a given program to see if the baseline analytic result is robust. Finally, probabilistic sensitivity analysis can simultaneously evaluate the effects of uncertainty for many input parameters. Sensitivity analysis is vitally important in economic evaluation precisely because so many assumptions are required. Many analysts, however, report only rudimentary sensitivity analyses. A common practice is to report one-way (varying a single parameter) sensitivity analyses for a highly selected group of model parameters. More complex and comprehensive approaches to evaluating uncertainty are less common. It is often difficult, therefore, for a consumer of published analyses to know how much uncertainty is attached to a reported result (Elliot and Payne, 2005). A new and important development within the past few years has been the development of statistical methods to evaluate uncertainty in cost-effectiveness analysis. Although these methods are largely of value for analyses carried out within clinical trials, they are a methodological advance and are increasingly being employed in published analyses. Interpreting Published Analyses: From Internal to External Validity Economic evaluations of breast cancer screening have reported incremental costs per year of life saved ranging from $3,000 to $84,000, a huge range (in Mehrez and Gafni, 2009). Why is there such extreme variation in the results of economic evaluations, and how confident can we be when applying the results of such research in our own clinical or health care context? First, apparent differences in reported results may be accounted for by methodological differences. The perspective of the analysis, inclusion/omission of selected costs, chosen discount rate, target year of the analysis (inflation), time horizon of the analysis, the way in which clinical efficacy is modelled or expressed, and whether average or incremental cost-effectiveness ratios are reported may affect the reported results (Mehrez and Gafni, 2009; Torrance, 2008). In addition, consumers of the literature must keep in mind that economic evaluations consider extremely specific clinical strategies, for specific populations, within a given national health care system. The comparison among screening mammography to no screening and to regular breast self-examination yield wildly different cost effectiveness ratios. Different screening intervals, high and low risk populations, young or old populations, and the efficiency of the health care system may dramatically affect cost-effectiveness ratios (Briggs, 2006). Thus the usual concerns about the generalizability of published clinical research to one’s own setting apply to economic research as well, but perhaps with additional force, because generalizability depends not only on clinical but on economic similarity between study and target populations. The economic attractiveness of a clinical strategy cannot easily be generalized to alternate populations, alternate health care systems, or alternate interventions. Interpreting Published Analyses: Decision Criteria and the Use of League Tables When one therapy costs more than an alternative and offers the potential for clinical harm, the alternative is clearly preferred. When a new therapy is cost saving and produces net health benefit, the program ought to be implemented. However, the works of Torrance and Feeny (2009) would indicate that new treatments usually produce a net health increase at an increase in net cost. This situation is more difficult to interpret. Whether the new treatment should be implemented depends on how much those who pay for health (society in general, subscribers to an HMO) are willing to pay for standardized units of health. This is not a question of fact but of value or preference. The traditional economic approach to this question has been to assume that resource allocation takes place within the context of a “fixed budget”. One simply ranks programs by their cost effectiveness and stops funding programs when the budget is fully allocated. The values of those who fund health care are implicitly expressed in the size of the budget allocated. Efforts to make this approach work in practice have foundered on the complexity of health care, the lack of availability of high quality analyses, and the limitations of current measurement approaches (Detsky and Nagle, 2006) The more common approach to determining what is worth paying for has been to examine the economic attractiveness of the intervention one is considering within the context of interventions that society is paying for at present (e.g., centre haemodialysis, hypertension treatment, and coronary revascularization). The use of “benchmarks” is gaining popularity. For example, Mehrez and Gafni (2009) suggested that programs that produce $20,000/QALY, $20,000 to $100,000/QALY and $100,000/QALY be rated as “strong evidence for adoption,” “moderate evidence for adoption,” and “weak evidence for adoption,” respectively. These thresholds are by no means uncontroversial or universally accepted. Economic Evaluation: Summary and Caveats Economic evaluation is a useful, increasingly widely used method to help clinicians and administrators formulate rational health policies. If economic evaluation is to be valuable for policy formulation, however, its limitations must be understood. First, economic evaluation is not prescriptive but, rather, a first (albeit important) step in the evaluation of an intervention. The distribution of costs and health benefits (who gains and who loses) and availability (matching resources to locations where they are accessible to those who require them) must also be considered. An additional reason the results of efficiency evaluation should not be blindly applied is that the methodology itself involves a standard theoretic framework that does not always validly represent the values of individuals or society. Critics have charged that economic evaluation does a poor job of reflecting society’s views on medical necessity, the moral injunction to aid those who are suffering (the “rule of rescue”), distributional issues, atypical preferences regarding health outcomes, attitudes toward risk, and so on. Economic evaluation, despite its limitations, remains an invaluable tool for understanding the economic and health consequences of clinical policies. References: Drummond, M.J.,, Sculpher, M. Torrance, G., O’Brien, B. and Stoddart G. (2005). Methods for the Evaluation of Health Care Programmes. US: Oxford University Press. Elliot, R. and Payne, K. (2005). Essentials of Economic Evaluation in Health Care. US: McGraw-Hill Publications. Zweifel, P., Breuer, F. And Kifmann, M. (2009). Health Economics. London: Routledge. Ketley, D. (2009). The Sustainability and Spread of Organizational Change: Modernizing Healthcare. London: Blackwell Publications. Briggs, A. H. (2006). Handling Uncertainty when Performing Economic Evaluation of Healthcare Interventions. (Health Technology Assessment). US: Woodridge. Gold, M.R., Siegel, J.E., Russell, L.B., Weinstein, M.C. (2007). Cost-effectiveness in Health and Medicine. New York: Oxford University Press. Detsky, A.S., and Naglie, I.G. (2006). A clinician’s guide to cost-effectiveness analysis. Ann. Intern. Med. 113:147. Rice, D.P., Hodgson, T.A., Kopstein, A.N. (2005). The economic costs of illness: a replication and update. Health Care Financ. Rev. 7:61. Briggs, A., Sculpher, M., Buxton, M. (2008). Uncertainty in the economic evaluation of health care technologies: the role of sensitivity analysis. Health Econ. 3:95. Torrance, G.W. and Feeny, D. (2009): Utilities and quality-adjusted life years. Int. J. Technol. Assess. Health Care 5:559. Shepard, D.S. and Weinstein, M.C (2005). Utility functions for life years and health status. Operations Res. 28:206. Mehrez, A. and Gafni, A (2009). Quality adjusted life years, utility theory, and healthy year equivalents. Med. Decis. Making 9:142. Torrance, G.W. (2008). Utility approach to measuring health-related quality of life. J. Chronic Dis. 40:593. Read More