Related Experiment Video
Updated: Oct 4, 2025
08:12
Fully Endoscopic Mitral Valve Repair with Percutaneous Cannulation of Groin Vessels
Published on: May 26, 2023
1.8K
Reevaluating Congenital Heart Surgery Center Performance Using Operative Mortality
Sharon-Lise T Normand1, Katya Zelevinsky2, Meena Nathan3
1Department of Health Care Policy, Harvard Medical School, Boston, Massachusetts; Department of Biostatistics, Harvard Chan School of Public Health, Boston, Massachusetts.
The Annals of Thoracic Surgery
|February 5, 2022
Summary
New methods for analyzing congenital heart surgery outcomes show improved performance categorization for many centers. These advanced approaches offer more precise quality reports for congenital heart centers.
Area of Science:
- Cardiovascular Surgery
- Health Services Research
- Biostatistics
Background:
- The Society of Thoracic Surgeons Congenital Heart Surgery Database (STS-CHSD) provides observed-to-expected (O/E) operative mortality ratios to North American congenital heart centers.
- Current methods for estimating O/E ratios may not fully capture center performance variations.
Purpose of the Study:
- To compare the current STS-CHSD approach for estimating O/E ratios with novel methods.
- To evaluate approaches that incorporate diagnosis-procedure interactions, additional risk factors, and refined confidence interval construction for characterizing center performance.
Main Methods:
- Bayesian additive regression trees (BART) and lasso models were used to link operative mortality to diagnosis-procedure categories, procedure-specific risk factors, and syndromes/abnormalities.
- Bootstrapping was employed to account for variation in the STS-CHSD (STS bootstrap) and lasso confidence intervals (CIs).
- O/E estimates, CI widths, and concordance of center performance categorizations were compared across methods.
Main Results:
- In an analysis of 110 centers and 98,822 operations, BART and lasso models yielded O/E ratios with greater variability and narrower CIs compared to the STS-CHSD.
- Concordance of performance categorization ranged from 84% (lasso) to 91% (STS Bootstrap), with over 70% of discordant centers showing improved categories.
- Discordant centers were characterized by smaller surgical volumes, fewer mortalities, and a higher proportion of patients with congenital lung abnormalities.
Conclusions:
- New analytical approaches can lead to significant shifts in hospital performance categories, with most centers improving their standing.
- Incorporating additional risk factors and addressing unaddressed variation is crucial for accurate quality reporting in congenital heart surgery.
- These findings highlight the importance of refining risk adjustment models for evaluating and comparing congenital heart center performance.
