2. Intervention Cost–Effectiveness: Overview of Main Messages

A frequent, often justified, criticism of cost-effectiveness analyses is that they address only one of many criteria that could be used to evaluate health interventions. Epidemiological, medical, political, ethical, and cultural factors often also play important roles in the decision to allocate resources to a specific health condition or intervention; however, determining how one might weigh cost-effectiveness ratios alongside these other considerations when setting priorities for spending is difficult. Musgrove (1999) shows how to take some of these connections into account, including circumstances in which cost-effectiveness is an adequate criterion by itself. One approach is for the policy maker to think of cost-effectiveness ratios as the relative "price" of purchasing a unit of health (a DALY, for instance) using different interventions. These costs, along with the budget constraint, can help determine the optimal allocation of resources among a given set of interventions.

Consider, for instance, a policy maker in a country in Sub-Saharan Africa facing the choice between treating Parkinson's disease and expanding malaria treatment programs while constrained by a fixed budget allocated by the ministry of finance. The cost per DALY averted of treating Parkinson's disease using carbidopa is vastly greater than that for the malaria control program. A simplistic interpretation of the cost-effectiveness information would be to expand the malaria treatment program to the maximum extent possible before turning to the treatment of Parkinson's disease. This solution could be desirable in some situations, particularly if the budget is large enough to deal fully with the malaria problem and still allow for some treatment of Parkinson's. Emphasizing as complete coverage as possible of a particular problem may be especially appropriate in an epidemic situation, in which turning to another disease first can mean the epidemic will be worse in the future. Devoting some resources to each problem instead of concentrating on either may be more sensible when neither presents the threat of a growing epidemic.

Asking policy makers to make a binary choice between two sets of interventions on the basis of cost-effectiveness ratios alone may be unrealistic and misleading. Rather, policy makers should first determine their willingness to trade off health improvements in children (malaria) versus the elderly (Parkinson's). Policy makers may want to avert at least some burden from Parkinson's even if these cases are relatively expensive to treat for each unit of health gained because of such considerations as the target age group, the socioeconomic status of target populations (including the extent to which they can obtain treatment from their own resources), and the ministry of health's ability to deliver the program effectively. After the tradeoffs have been made and can be represented by an indifference curve (see the figure and explanation), the cost-effectiveness information is useful in determining how much of the policy makers' fixed budget should be allocated to each intervention—that is, at what coverage of one problem should they start devoting resources to the other? The indifference curve represents health planners' willingness to trade off between investment in antimalarial drugs and treatment for Parkinson's based on all the relevant factors and independent of the budget constraint.

The solid line represents the budget, and its slope is the ratio of the cost-effectiveness ratios of the two interventions. The dashed line represents an alternative scenario in which the cost-effectiveness of treating Parkinson's is better (more DALYs can be gained) than the ratio represented by the solid line. The axes show how many DALYs can be gained from each treatment, so a large number of DALYs corresponds to a low "price" of health or cost per DALY. The figure shows the simple case, in which these prices are constant for either budget line—that is, expanding either program does not raise the unit cost—although this case is unlikely when part of the population is difficult to reach, is harder to treat, or has more severe disease, in which case the rise in unit cost means that an intervention becomes relatively less cost-effective, giving a further reason to start devoting resources to an alternative intervention.

When the price of buying a unit of health to treat Parkinson's is relatively high in terms of cost per DALY averted, the relatively flat (solid) budget line applies, and the optimal balance of investment in the two interventions is at point X. If the cost of buying a unit of health to treat Parkinson's is relatively low, then the steeper (dashed) budget line applies and the relative allocation of resources is represented by point Y. Therefore, policy makers would allocate relatively more resources to treating Parkinson's when the price of buying a unit of health through this intervention is relatively low, and they would allocate fewer resources when the price of health obtained through this intervention is relatively high. The figure shows the general likely shape of an indifference curve, but one possibility is that policy makers' willingness to trade off between buying health from the two approaches is just a straight line, in which case they would want to invest the maximum amount possible in the lower-cost intervention (malaria) before turning to the higher-cost intervention (Parkinson's). The role of the cost-effectiveness information is to make policy makers aware of differences in the price of improving health using different interventions. Interventions with a high price should, all else being equal, be used less, whereas those with a low price should be used to a greater extent.