Adequate Use of Economic Evaluations Based on the Endpoints of a Registrative Clinical Trial: Cost-Minimization Analysis and the Return of the Living Dead

Abstract

Cost-effectiveness analysis (CEA) has been widely applied to compare alternative options when a difference in health effects is observed. In contrast, a cost-minimization analysis (CMA) must be performed if the health effects of the alternatives can be considered reasonably similar. However, several authors have been suggested the death of CMA and supported the use of CEA alongside registrative clinical trials, even if the clinical end points of a study fail to demonstrate a statistically significant difference. The aim of the article is to discuss the potentially misleading implications of the inadequate use of CEA in translating results into clinical recommendations. A case study based on main findings from a published CEA evaluating the use of direct acting oral anticoagulants (DOACs), vitamin K antagonists, or antiplatelet drugs for the prevention of stroke in patients with atrial fibrillation, has been considered. The CEA recommends the use of apixaban as the first choice among DOACs, since it ranks the highest on the balance of efficacy, safety, and cost. However, no clinical evidence supporting this recommendation is available. In contrast, from a therapeutic perspective, several aspects support the preferential use of the other DOACs as a better first choice. In the case-study discussion, the step in which the incremental net benefit is calculated in the absence of at least one better alternative option either in regard to QALYs or total cost is critical and may promote incorrect conclusions and misleading clinical recommendations. The article proposed synoptic framework of the adequate use of economic evaluations based on the endpoints of registrative clinical trials, considering limitations on the use of CEAs and the expansion of CMA applications. Thus, CMAs should not only be limited to comparisons of products with evidence of efficacy from studies using an equivalence hypothesis test.