The effect of hospital volume on resection margins in rectal cancer surgery

Background We hypothesized that after controlling for case-mix differences, the rates of positive resection margin after rectal cancer surgery vary substantially in the United States and that high-volume hospitals have lower margin positivity rates. Materials and methods Patients treated with oncologic resection for stage I-III rectal cancer were selected from the 1998-2010 National Cancer Data Base. Hierarchical regression models were used to calculate risk- and reliability-adjusted positive margin rates and hospital level variability in positive margin rates using Empirical Bayes techniques. Results A total of 113,113 patients were treated at 1446 hospitals. The mean overall risk- and reliability-adjusted positive margin rate was 7.3%. High-volume hospitals did not have a lower rate of adjusted margin positivity (7.4%, P = 0.75). When both case mix and hospital volume differences were factored into the model, variability in margin positivity rates increased by 9.8%, implying that referral to high-volume hospitals alone would not improve margin positivity rates. Conclusions Rectal cancer margin positivity rates vary substantially in the United States, despite adjusting for differences in case mix. These results support standardization of surgical technique and pathologic assessment as part of a broader initiative that identifies and refers patients to higher performing hospitals rather than simply to higher volume hospitals.