Will a Conservative Case Selection Strategy Improve Hospital-Level TAVR Performance Metrics?

BACKGROUND: Current national performance metrics rank transcatheter aortic valve replacement (TAVR) centers based on risk-adjusted outcomes. This could make operators/centers less inclined to offer TAVR in high-risk cases. METHODS: We used simulation models based on registry data to explore whether avoiding high-risk TAVR cases would improve the hospitals' comparative risk-adjusted TAVR outcomes. This multicenter, retrospective cohort study included all adults (≥18 years) who underwent TAVR in the Society of Thoracic Surgeons/American College of Cardiology Transcatheter Valve Therapy registry in 2021. We evaluated hospital-level, risk-adjusted outcomes, before and after simulating the omission of the top 10% highest risk patients. The primary outcome was a 30-day composite of death, stroke, VARC major/life-threatening/disabling bleeding, renal failure, or moderate/severe para-valvular leak. The secondary outcome was 30-day death. We used the mean difference±SD in the win ratio and observed/expected ratio for evaluation of the primary and secondary outcomes, respectively. RESULTS: There were 43 907 TAVR cases with available primary outcome data and 56 982 cases with available secondary outcome data. Median age was 79 (73-84) years, 57% were men, and 93% were White race. Our simulation demonstrates that, on average, excluding the top 10% highest risk patients from centers' case mix would not change their hospital-level, risk-adjusted win ratio (mean difference, 0.002±0.067; P=0.60) or observed/expected ratio (mean difference, 0.003±0.633; P=0.90). CONCLUSIONS: Hospital-level, risk-adjusted TAVR outcomes did not consistently improve with simulated strategies of avoiding the highest-risk cases. Operators and centers can be reassured that they can continue to offer TAVR to high-risk patients, as clinically indicated, without the sole focus on being penalized via quality measures.

Duke Scholars

Author Sreekanth Vemulapalli Medicine, Cardiology

Author Andrzej Stanislaw Kosinski Biostatistics & Bioinformatics, Division of Biostatistics