Cardiovascular Perspectives Cost, Effectiveness, and Cost-Effectiveness

Incremental or marginal cost-effectiveness ratios are founded on a number of assumptions that weaken their suitability as a way to balance competing economic and clinical priorities. We therefore propose a more relevant and responsible way to gauge the impact of clinical management strategies based on “consumer protection” principles— explicit disclosure of (1) the total magnitude of expected benefit in the target population (in life-years or qualityadjusted life-years), (2) the total monetary cost (in per capita inflation-adjusted dollars), and (3) a formal plan by which the added costs would be paid. Health policy decisions should therefore be based on the inherent tradeoffs in the component measures of cost and effectiveness and not on a simple ratio of the two. Cost-effectiveness ratios are thereby rendered superfluous. Incremental or marginal cost-effectiveness ratios are widely viewed as rational ways to balance competing clinical and economic priorities that arise as a consequence of the inevitable disconnect between an individual’s wants and the society’s willingness to pay for those wants.1–5 The purpose of this essay is to question the practical relevance of these ratios with respect to strategic planning in health care, and offer a suitable alternative. In doing so, we will focus on the more pragmatic issues underlying cost-effectiveness analysis, and gloss over a variety of technical details such as the difference between a privately financed free market and the publicly financed healthcare market, between costs and charges, and between unadjusted and quality-adjusted outcomes. Our pragmatic perspective is not likely to be welcomed by orthodox experts in cost-effectiveness analysis. We believe, however, that any such criticisms can be blunted by recognizing the important distinctions between regulatory and clinical decisions. As clinicians writing to a clinical audience, we will emphasize the role of cost-effectiveness in clinical decision making throughout much of our discussion, but we will comment on its regulatory role as well. At its core, cost-effectiveness analysis is no less than a utilitarian moral calculus—one that quantifies the value of any action (even the plucking of a goose) as the ratio of the observable bad qualities, the costs, to the observable good qualities, the benefits. For any action, the lower the ratio of cost to benefit (of hissing to feathers), the higher the value of that action. Although we take no issue with the sophisticated methods used to compute the component values of cost and effectiveness, we will show that the conventional costeffectiveness ratio suffers from a number of shortcomings as a summary statistic and decision criterion. We are all familiar with such measures. The price-earnings ratio, for example, values a company in terms of the street price of its stock in proportion to its actual or projected earnings per share. Similarly, the lipoprotein ratio quantifies cardiovascular risk in terms of the ratio of (bad) low-density lipoprotein cholesterol to (good) high-density lipoprotein cholesterol. In the typical healthcare application, the amount of “bad” associated with any action is usually quantified in terms of monetary costs, and the amount of “good” is quantified in terms of the savings in lives, life-years, or quality-adjusted life-years. The conventional threshold of cost-effectiveness is most often taken to be $50 000 per quality-adjusted life-year (just as the conventional threshold for statistical significance is 0.05). Actions valued below this threshold are commonly considered “cost-effective” and those above are not.6–23 The justification for this singular boundary dates back to a 1980 report on Medicare coverage for treatment of end-stage renal disease, which projected the number of such cases would stabilize at approximately 90 000 patients in 1995 at an inflation-adjusted cost of $4.5 billion.24 This translates to a ratio of $50 000 per life-year (unadjusted for quality). In fact, the numbers have turned out to be a little off. Actual enrollment for 1995 almost tripled (251 214 patients) and the inflation-adjusted cost almost doubled ($8.8 billion)— equivalent to a ratio of only $35 154 per life-year. In any event, numeric benchmarks such as this are a gross oversimplification of a highly complex process. All assessments of value are inherently subjective, and the best way to gauge the value of alternative management strategies is by an up-to-date, real-world, local comparison of the multiple factors entering into the interaction between cost and effectiveness. Nevertheless, if we choose to short-circuit this process so as to make simple categorical judgments regarding cost-effectiveness—the canonical goal of cost-effectiveness