Wide variation in mitral valve repair rates among US surgeons: Analysis of Medicare claims data

Hanghang Wang,Chen Dun,Martin A Makary,Christi Walsh,Yi Fan,Emily Rodriguez,Deven Patel,Alice Zhou,Armaan Akbar,Glenn Whitman,James S Gammie

DOI: https://doi.org/10.1016/j.jtcvs.2024.10.035

2024-10-25

Abstract:Objective: Mitral valve repair is the preferred treatment for primary mitral regurgitation and offers significant short- and long-term advantages over valve replacement. This study was designed to evaluate the contemporary national mitral valve surgery practice patterns, focusing on the impact of surgeon-specific factors, such as operative volume and years of practice, on repair rates. Methods: A retrospective analysis was conducted using 100% Medicare fee-for-service claims data over a 3-year period (January 2020 to December 2022). Mitral valve procedures were identified using specific Current Procedural Terminology codes. We excluded patients with active infective endocarditis, mitral stenosis, or a history of previous mitral valve repair or replacement. Multivariable binomial regression was used to assess the impact of surgeon-specific factors on repair rates. Results: We identified 2072 surgeons in 770 hospitals who performed 12,339 mitral valve operations, with an overall repair rate of 68.8%. The median number of mitral valve operations performed per surgeon during the 3-year study period was 3 (interquartile range [IQR], 2-7), and the median number of mitral valve repairs was 2 (IQR, 1-5). A subset of 312 surgeons (15%) performed more than 10 mitral valve procedures each and more than one half (57%) of all repairs nationally. This subgroup's median repair rate was 77%, with significant variability within the group: 17% of surgeons had a repair rate below 50%, 59% had a repair rate between 50% and 90%, and 24% had a repair rate greater than 90%. Multivariable regression analysis indicated significant associations between repair rates and surgeon-specific factors, including surgical volume, years of practice, and region of practice. Each additional procedure was associated with a 1.5% average increase in repair rate likelihood (95% confidence interval, 1.2%-1.8%, P < .001), and each additional year of practice was associated with a 1.4% average increase (95% confidence interval, 0.8%-2%, P < .001). Regional differences were notable: surgeons in the South demonstrated lower repair rates (median 71%; IQR, 55%-85%) compared with those in the Northeast (median, 78%; IQR, 68%-91%, P = .02) and Midwest (median 86%; IQR, 63%-92%, P = .04). Conclusions: This study has identified significant variability in mitral valve repair rates among surgeons who treat Medicare beneficiaries. Notably, even among the surgeons responsible for most of these procedures, the variability in repair rates is pronounced. These findings suggest substantial opportunities to improve outcomes for patients undergoing mitral valve operations in North America.