Variation in Drain Management After Pancreatoduodenectomy Early Versus Delayed Removal

Objective: The objectives are to report practice patterns and management of operatively placed drains and to compare outcomes in patients with early versus delayed drain removal after pancreatoduodenectomy. Background: Early drain removal after pancreatoduodenectomy, when guided by postoperative day (POD) 1 drain fluid amylase (DFA-1), is associated with reduced rates of clinically relevant postoperative pancreatic fistula (CR-POPF). However, whether surgeons have altered their management based on this strategy is unknown. Methods: The American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) 2014 Participant Use File was queried to identify patients having undergone pancreatoduodenectomy (n = 3093). Patients with intraoperatively placed drains were stratified according to measurement of DFA-1 and day of drain removal. Patients with POD 1 DFA-1 of <= 5000 U/L whose drains were removed by POD 3 were propensity score-matched with patients whose drains were removed after POD 3. Results: Of 2698 patients, 580 (21.5%) had a DFA-1 recorded. Measurement of DFA-1 was associated with earlier time to drain removal and shorter postoperative length of stay (P < 0.01). Propensity score matching revealed that early drain removal when DFA-1 was <= 5000 U/L was associated with significant (P < 0.05) reductions in overall morbidity (35.3% vs 52.3%), CR-POPF (0.9% vs 7.9%), and length of stay (6 vs 8 days). Conclusions: Significant variation exists in the use of drain fluid amylase in the management and timing of surgical drain removal after pancreatoduodenectomy. Clinical outcomes are best when drain fluid amylase is low and operatively placed drains are removed by POD 3.