Hospital and surgeon surgical valvar volume and survival after multivalve cardiac surgery in Medicare beneficiaries

Kyle A McCullough; John B Eisenga; Jasjit K Banwait; Alessandro Gasparini; Katherine B Harrington; Timothy J George; Kelley A Hutcheson; Robert L Smith 2nd; William T Brinkman; J Michael DiMaio; Justin M Schaffer

doi:10.1016/j.jtcvs.2024.08.044

Hospital and surgeon surgical valvar volume and survival after multivalve cardiac surgery in Medicare beneficiaries

J Thorac Cardiovasc Surg. 2024 Sep 5:S0022-5223(24)00778-5. doi: 10.1016/j.jtcvs.2024.08.044. Online ahead of print.

Affiliations

¹ Department of Cardiovascular Research, Baylor Scott & White Research Institute, Plano, Tex. Electronic address: Kyle.McCullough@BSWHealth.org.
² Department of Cardiovascular Research, Baylor Scott & White Research Institute, Plano, Tex.
³ Red Door Analytics AB, Stockholm, Sweden.
⁴ Department of Cardiac Surgery, Baylor Scott & White The Heart Hospital Plano, Plano, Tex.
⁵ Department of Cardiovascular Research, Baylor Scott & White Research Institute, Plano, Tex; Department of Cardiac Surgery, Baylor Scott & White The Heart Hospital Plano, Plano, Tex; Department of Biomedical Engineering, Texas A&M University, College Station, Tex.

PMID: 39243959
DOI: 10.1016/j.jtcvs.2024.08.044

Abstract

Objective: Long-term outcomes after multivalve cardiac surgery remain underevaluated.

Methods: Medicare administrative claims from 2008 to 2019 identified beneficiaries undergoing multivalve surgery. Operative characteristics were doubly adjudicated using International Classification of Diseases and Current Procedural Technology codes. A multivariable flexible parametric model evaluated predictors of survival; regression standardization was performed to predict standardized survival probabilities (SSPs) at varying percentiles of annual valvar volume.

Results: Of 476,092 cardiac surgeries involving the aortic (AVS), mitral (MVS), or tricuspid (TVS) valve, 63,083 (13.3%) were identified as involving multi-valve surgery, including 22,884 MVS/TVS, 30,697 AVS/MVS, 3443 AVS/TVS, and 6059 AVS/MVS/TVS. Surgery occurred at 1157 hospitals by 2922 surgeons. Annual valvar volume (total AVS/MVS/TVS) was tallied for surgeons and hospitals. Median survival varied substantially by type of multivalve surgery: 8.09 (95% CI: 7.90-8.24) years in MVS/TVS, 6.65 (95% CI: 6.49-6.81) years in AVS/MVS, 5.77 (95% CI: 5.37-6.13) years in AVS/TVS, and 6.02 (95% CI: 5.64-6.38) years in AVS/MVS/TVS. SSPs were calculated across combined hospital/surgeon volume percentiles; the median SSP increased with increasing percentile of combined hospital/surgeon volume: 5th percentile, 5.77 (95% CI: 5.58-5.98), 25th percentile, 6.18 (95% CI: 6.07-6.28), 50th percentile, 6.56 (95% CI: 6.44-6.68), 75%th percentile, 6.86 (95% CI: 6.75-6.97), and 95th percentile, 7.58 (95% CI: 7.34-7.83) years, respectively.

Conclusions: Survival varied significantly by type of multivalve surgery, worsened with addition of concomitant interventions and improved substantially with increasing annual hospital and surgeon volume. Hospital volume was associated with an improved early hazard for death that abated beyond 3 months post surgery, while surgeon volume was associated with an improved hazard for death that persisted even beyond the first postoperative year. Consideration should be given to referring multivalve cases to high-volume hospitals and surgeons.

Keywords: Medicare; aortic valve; mitral valve; multivalve surgery; survival; tricuspid valve; volume-outcome relationship.