Title: Population-level cost-effectiveness analysis: The individual net benefit from a causal perspective
Authors: Andrew Spieker - Vanderbilt University Medical Center (United States) [presenting]
Jason Roy - Rutgers University (United States)
Nandita Mitra - University of Pennsylvania (United States)
Abstract: Health policy decisions regarding a particular treatment or intervention are generally made on the basis of aggregate information on both clinical effectiveness and cost. The net monetary benefit has been used as a measure to evaluate the comparative cost-effectiveness of two interventions. Briefly, this measure seeks to quantify the extent to which a treatment's level of efficacy justifies its associated cost, on average, at some willingness-to-pay threshold. Complexities of observational cost data including confounding, censoring, truncation by death, and time-dependent treatment can all impede causal estimation of cost-effectiveness. We present a marginal structural model framework for the net monetary benefit in order to overcome these challenges, thereby enabling causal interpretation and promoting refined cost-effectiveness comparison of clinically defined subgroups. We additionally present a novel visualization tool for cost-effectiveness that overcomes limitations of the current widely used cost-effectiveness acceptability curve.