Pacific C
A Second-Generation Cost-Effectiveness Acceptability Curve Based on a Bayesian Credible Interval for the Net Monetary Benefit (306503)
Jeffrey Blume, Vanderbilt University Medical Center*Andrew Justin Spieker, Vanderbilt University Medical Center
Keywords: Bayesian, Net Monetary Benefit, Cost-Effectiveness, Visualization
In an era of rising healthcare spending, policy makers must assess both an intervention’s clinical effectiveness and cost. Data sources with patient-level cost outcomes are becoming available, broadening the range of clinical questions that can be answered and the statistical methods that can be applied. The widely used acceptability curve quantifies strength of evidence in favor of cost-effectiveness, but fails to distinguish between lack of evidence regarding cost-effectiveness and strong evidence of near-null effects. We present a visualization tool for cost-effectiveness comparisons using the second-generation p-value (SGPV) based on a Bayesian credible interval. The SGPV measures the overlap between the credible interval for net monetary benefit and a pre-specified interval-null hypothesis, and serves as a useful tool for decision-making. We illustrate this approach by comparing cost-effectiveness of competing adjuvant therapies in a SEER-Medicare linked database of endometrial cancer patients. Appropriate visualization tools allow us to navigate the realities of modern healthcare and emerging cost data in order to better inform future policy decisions.