What's all this about pharmacoeconomics?

Business Economics, April, 1995 by Richard J. Willke

These are boom times for pharmacoeconomics. Pharmaceutical companies, consulting firms, academic departments in various disciplines, insurance companies and government agencies all over the world have jumped in with both feet. Pharmacoeconomic studies are being done at an exponentially increasing rate, and the methodological sophistication of those studies is growing as well. The demand for pharmacoeconomic evidence about new and existing compounds is widespread. As a byproduct, guidelines on how pharmacoeconomics should be done are also proliferating. Economists are migrating into this new field; journals are offering special articles about the phenomenon.

What is it? Where has it come from, where is it going, and why all the fuss? At one level, the answer is simple economics: health care resources are limited and must be allocated efficiently. The dollars that can be spent on health care, both public and private, cannot cover the full range of services health care has to offer. One of the spotlights has been on drugs, and increased attention to economic evaluation of drugs has been the result. However, pharmacoeconomics is being practiced by people from many disciplines besides economics, e.g., pharmacy and medicine, and being consumed by many more. As a new and potentially influential field, there are concerns about both its production and its consumption. My purpose here is to provide an overview of its methodology and applications and to highlight some of the debates and concerns about this field, thereby helping the business economist - producer, consumer, or innocent bystander - to evaluate it.

A SHORT COURSE ON METHODOLOGY

Pharmacoeconomics, narrowly defined, is cost-benefit analysis of health care interventions involving ethical drugs. As such, its methods share a common heritage with any other microeconomic, empirical evaluations of interventions, be their purposes educational, labor-force related, or project development. More broadly defined, pharmacoeconomics includes cost-of-illness studies and quality of life studies, which will not be discussed here. Although industry level analyses, such as the cost of developing new drugs, are common topics in pharmacoeconomics, they are not considered to be per se part of this field.

Pharmacoeconomic cost-benefit analysis most frequently takes the form of a cost-effectiveness analysis. Outcomes of health care interventions are difficult to capture in dollar terms because a health improvement usually occurs, perhaps a death averted, which is difficult to value in dollar terms. Therefore, cost-effectiveness analysis captures the benefit of the intervention using some other outcome measure, such as heart attacks prevented, remissions achieved, life-years saved or "quality-adjusted" life-years saved. Costs measured should include not only the cost of the drug but also any medical or other "direct" (such as transportation) costs involved in the treatment "episode." The treatment episode may last just a few days, such as for some antibiotics, or the rest of the patient's life, in many chronic or disabling conditions. The intervention must be contrasted to a standard or reasonable alternative treatment, so that marginal changes in both costs and effects can be measured.

Given measures of the marginal costs and effects of the drug intervention, four main combinations of outcomes are possible. If the drug both improves outcomes and decreases overall costs of treatment, it is a clear winner, sometimes called a dominant treatment strategy. Two subcases here are: (1) when the drug does not change the outcome but it reduces costs -- it is then a cost-minimizing treatment strategy; or when it improves outcome with no change in cost (there's no particular term for this). Both of these results would also be favorable to use of the drug. Secondly, if the drug is less effective and more costly than the alternative treatment, it is just as clearly a bad choice. The third situation occurs when the drug both improves outcomes and increases costs. Here, the choice is less clear and depends on the cost-effectiveness ratio:

Cost-effectiveness = change in costs due to treatment/change in outcome due to treatment

The smaller the cost-effectiveness ratio, the more cost-effective the drug is considered to be. For example, if a drug costs $1,000 per life-year saved, it would normally be considered very cost effective. If it costs $100,000 per life-year saved, it is not very cost effective but, at present, not clearly out of range. One of the problems in this field has been making judgments about what acceptable cost-effectiveness ratios are; we'll return to this later.

Finally, it is also possible that the drug both reduces cost and worsens outcomes. One may construct a cost-effectiveness ratio here, but its interpretation is quite different - "cost saved per life-year lost." When compared to standard treatment, a drug that significantly worsens outcomes may or may not be considered for use depending on the condition and the source of the cost savings. A less expensive, less effective antibiotic for ear infections may be a reasonable choice under some financial conditions; a drug that saves total hospital costs because people using it are more likely to die (resulting in shorter hospital stays) than under standard treatment would not be considered ethical to use.