Background: To better understand re-excision practice patterns after breast-conserving therapy (BCT), we evaluated variation in surgeon-specific re-excision rates and associated factors. Study Design: We performed a retrospective analysis using Medicare claims from 2012 to 2018 to identify patients undergoing BCT and subsequent breast resection procedures within 12 months. We compared rates before and after the 2014 “no tumor on ink” consensus guideline. A hierarchical logistic regression model was also used to evaluate patient and physician characteristics associated with re-excision. Results: We identified 291,065 female Medicare beneficiaries who underwent an initial BCT procedure, of which 19.0% had a re-excision. The overall re-excision rate was 22.1% in the pre-guideline period and 17.2% in the post-guideline period. For the 5,337 physicians that performed more than 10 initial BCT procedures during the study period, their physician-level re-excision rate ranged from 0% to 91.7% (median 18.2%). In total, 17.5% of the physicians had a re-excision rate greater than the expert consensus cutoff of 30%. The percentage of outlier physicians decreased from 22.2% in 2012 to 8.8% in 2017. High surgeon volume of BCT was associated with a lower re-excision odds (≥51 cases vs ≤20 cases: adjusted odds ratio 0.78; 95% CI 0.74 to 0.82; 21 to 50 cases vs ≤20 cases: adjusted odds ratio 0.92; 95% CI 0.88 to 0.96). Patient factors associated with decreased odds of re-excision were age older than 75 years and Northeast region of the US (adjusted odds ratio 0.93; 95% CI 0.89 to 0.98). Conclusions: Marked variation exists in surgeon re-excision rates among patients undergoing BCT, which might represent unnecessary operations for patients and a financial burden to the healthcare system. Formalizing a re-excision frequency metric could have implications for quality improvement and data-driven surgeon feedback aimed at reducing unwarranted variation.
ASJC Scopus subject areas