A Statistical Method for Measuring Cost-Effectiveness Ratio

Economic evaluation of competing health care programs or interventions involves Cost-Effectiveness Analysis. In general, the cost effectiveness ratio of a test intervention or program is defined as the ratio of its incremental cost, over its best available alternative or referent intervention, to the incremental health benefit that it produces. However, such measures could produce invalid results in the presence of interactions between the cost and other independent variables. In this paper we revisit the original concept of assessing cost effectiveness of health interventions through regression models. We demonstrate how this method could be applied to situations where interactions exist between the cost and other independent variables. It is important to note that we are considering a situation where a single health care intervention or treatment in which survival time is the primary clinical outcome of interest and cost and effects are collected at the level of each patient for an intervention and a control group. The proposed method allows for varying level of censorship among different groups of patients at different cost levels. It is also assumed that the cost of treatment takes only finitely many values. The asymptotic normality of the estimators is established. In addition, the advantages of this method over some of the existing methods are discussed.