Cost-utility analysis of TAVI compared with surgery in patients with severe aortic stenosis at low risk of surgical mortality in the Netherlands

Abstract

BackgroundThere is growing evidence to support the benefits of transcatheter aortic valve implantation (TAVI) over surgical aortic valve replacement (SAVR) in patients with symptomatic severe aortic stenosis (sSAS) who are at high- or intermediate-risk of surgical mortality. The PARTNER 3 trial showed clinical benefits with SAPIEN 3 TAVI compared with SAVR in patients at low risk of surgical mortality. Whether TAVI is also cost-effective compared with SAVR for low-risk patients in the Dutch healthcare system remains uncertain. This article presents an analysis using PARTNER 3 outcomes and costs data from the Netherlands to inform a cost-utility model and examine cost implications of TAVI over SAVR in a Dutch low-risk population.MethodsA two-stage cost-utility analysis was performed using a published and validated health economic model based on adverse events with both TAVI and SAVR interventions from a published randomized low risk trial dataset, and a Markov model that captured lifetime healthcare costs and patient outcomes post-intervention. The model was adapted using Netherlands-specific cost data to assess the cost-effectiveness of TAVI and SAVR. Uncertainty was addressed using deterministic and probabilistic sensitivity analyses.ResultsTAVI generated 0.89 additional quality-adjusted life years (QALYs) at a €4742 increase in costs per patient compared with SAVR over a lifetime time horizon, representing an incremental cost-effectiveness ratio (ICER) of €5346 per QALY gained. Sensitivity analyses confirm robust results, with TAVI remaining cost-effective across several sensitivity analyses.ConclusionsBased on the model results, compared with SAVR, TAVI with SAPIEN 3 appears cost-effective for the treatment of Dutch patients with sSAS who are at low risk of surgical mortality. Qualitative data suggest broader societal benefits are likely and these findings could be used to optimize appropriate intervention selection for this patient population.