Measuring the Cost-Effectiveness of Cancer Care

June 1, 1996

This timely and informative review describes the components of a cost-effectiveness analysis and provides useful commentary on various ways to measure them. It may be helpful, however, to take a step back and compare cost-effectiveness analysis to the other basic approaches to economic analysis.

This timely and informative review describes the components ofa cost-effectiveness analysis and provides useful commentary onvarious ways to measure them. It may be helpful, however, to takea step back and compare cost-effectiveness analysis to the otherbasic approaches to economic analysis.

In all health economic analyses, the goal is to compare one treatmentor strategy with another. The real question is not how much itcosts to deliver a particular treatment, but how much more itcosts to provide that treatment than the most reasonable alternative.Sometimes the alternative is a "no treatment" strategy,but this isn't the equivalent of a "no cost" strategy.For example, the real difference in direct medical costs betweena strategy of giving adjuvant chemotherapy for a particular diseaseand a no treatment alternative is not limited to the cost of thechemotherapeutic agents, antiemetics, and nurse and physiciantime required to administer them. The cost of caring for the patientsin subsequent years is also a component of the direct medicalcost of each strategy. The additional or incremental cost of theadjuvant therapy strategy may actually be lower than the costof the treatment itself if patients in the no treatment groupexperience poorer outcomes and consequently consume more medicalresources in later years.

All economic analyses, then, examine the difference in cost betweenalternative strategies. They differ in how they measure the benefitsresulting from those strategies. As shown in Table 1, the fourbasic types of economic analysis measure these benefits in fourdifferent ways. In a cost-minimization study, the two treatmentsare assumed to produce comparable benefits. The less expensivetreatment is therefore the preferred one. But very few medicalinterventions produce truly equivalent outcomes. More commonly,the question is whether the additional benefit conferred by themore expensive treatment is sufficient to justify the additionalcost.

A cost-benefit analysis addresses this question by assigning adollar value to the health outcome. If the incremental benefitof one treatment over another, measured in monetary terms, isgreater than or equal to the incremental cost, then the additionalcost is considered justified. Cost-benefit analyses are used infrequentlyin evaluating health interventions because of the difficulty inherentin assigning a dollar value to a life saved and other health outcomes.

Cost-effectiveness analysis measures the benefit of a health-careintervention in units of medical effect. For example, the cost-effectivenessof a new chemotherapeutic agent in comparison with a cheaper alternativecould be assessed by calculating the additional cost (in dollars)per each additional complete response. However, one of the goalsof cost-effectiveness analysis is to guide decisions about whetherexpenditures for a particular intervention are justified comparedwith alternative uses of those limited resources, including treatmentsfor other diseases. This judgment is facilitated if health benefitsare measured in units that are common across diseases. Years oflife saved is the most frequently used measure. Cost-effectivenessratios are therefore usually expressed in units of dollars peryear of life saved.

How does one judge what constitutes a reasonable number of dollarsper year of life saved? Standards for this measure have been establishedby examining cost-effectiveness ratios for interventions thatare generally regarded by society as reasonable and those thatare generally regarded as inordinately expensive for the degreeof benefit produced. The conclusion is that interventions costingunder approximately $50,000 per year of life saved are cost-effective,and those costing over $100,000 are cost-ineffective. Cost-effectivenessratios falling between these levels are in a gray zone.

Cost-utility analysis--But medical interventions are notdirected exclusively at saving years of life. Much of the health-caredollar is spent on therapies that are given with palliative intent.Cost-utility analysis is a particular type of cost-effectivenessanalysis that takes into account the impact of a health interventionon quality of life as well as length of life. As described bySchulman and Yabroff, this is done by measuring quality-adjustedsurvival in the form of quality-adjusted life-years (QALYs). Theunits of a cost-utility ratio are thus dollars per QALY. Of allthe types of economic evaluation, cost-utility analyses assesshealth benefits in the most clinically meaningful way. They areappearing with increasing frequency in the clinical literatureand may become the dominant mode of economic analysis as user-friendlymethods of utility (preference) assessment become available.

Should the practicing clinician care about economic analyses ofhealth care intervention? One could ask whether it is appropriateto publish economic analyses in the clinical literature at all.In the United States, though not necessarily in all developedcountries, it is generally believed that the physician shouldfunction as the patient's advocate and in this role should offerany treatment likely to be of net benefit, regardless of the costto society. But some clinicians serve other roles as well, asadministrators and policy makers, jobs in which they need to considerwhether the costs of various health-care services are justifiedby the benefits they produce. And all are citizens and tax payers.If physicians are to have a meaningful voice in the debate overthe allocation of health- care dollars, they must understand thelanguage in which the discussions are being held.