OBJECTIVE:This multicenter study sought to prospectively evaluate a drain management protocol for pancreatoduodenectomy (PD). BACKGROUND:Recent evidence suggests value for both selective drain placement and early drain removal for PD. Both strategies have been associated with reduced rates of clinically relevant pancreatic fistula (CR-POPF)—the most common and morbid complication after PD. METHODS:The protocol was applied to 260 consecutive PDs performed at two institutions over 17 months. Risk for ISGPF CR-POPF was determined intraoperatively using the Fistula Risk Score (FRS); drains were omitted in negligible/low risk patients and drain fluid amylase (DFA) was measured on postoperative day 1 (POD 1) for moderate/high risk patients. Drains were removed early (POD 3) in patients with POD 1 DFA ≤5,000 U/L, whereas patients with POD 1 DFA >5,000 U/L were managed by clinical discretion. Outcomes were compared with a historical cohort (N = 557; 2011–2014). RESULTS:Fistula risk did not differ between cohorts (median FRS4 vs 4; P = 0.933). No CR-POPFs developed in the 70 (26.9%) negligible/low risk patients. Overall CR-POPF rates were significantly lower after protocol implementation (11.2 vs 20.6%, P = 0.001). The protocol cohort also demonstrated lower rates of severe complication, any complication, reoperation, and percutaneous drainage (all P < 0.05). These patients also experienced reduced hospital stay (median8 days vs 9 days, P = 0.001). There were no differences between cohorts in the frequency of bile or chyle leaks. CONCLUSIONS:Drains can be safely omitted for one-quarter of PDs. Drain amylase analysis identifies which moderate/high risk patients benefit from early drain removal. This data-driven, risk-stratified approach significantly decreases the occurrence of clinically relevant pancreatic fistula.