RESUMEN
BACKGROUND: The mainstay of treatment for pancreatic cancer is surgical resection; however, positive surgical margins remain commonplace. We identified hospitals with higher than predicted rates of positive margins and isolated factors that caused this discordance. METHODS: This is a retrospective review of patients with head of the pancreas adenocarcinoma in the National Cancer Database between 2004 and 2015. A nomogram was used to calculate the observed to expected positive margin rates (O/E) for facilities. If the O/E differed significantly (P < .05), it was considered an outlier. RESULTS: Among a total of 19 968 patients, 24.3% had positive margins. Among hospitals with lower than expected positive margin rates, 73.6% were academic or research programs, 17% were comprehensive community cancer programs, and none were community cancer programs (P = .0002). Within the group with higher than expected positive margin rates, 47% were comprehensive community cancer programs and 38.6% were academic or research programs (P = .0002). The mean hospital volume was higher in the low positive margin group (110.4 vs 48.8, P < .0001). CONCLUSIONS: Facility type and hospital volume can predict improvement in the O/E ratio for margin positivity in pancreatic adenocarcinoma resection. Surgeons should consider referral to academic or research facilities with higher case volumes for improved surgical resection.