ABSTRACT
OBJECTIVE: To validate the 2016 ISGPS definition and grading scheme and investigate whether it segregates into distinct subclasses. METHODS: A total of 522 patients undergone pancreaticoduodenectomy in two pancreatic centers were reviewed. The 2016 ISGPS scheme was validated by comparing clinical and economic outcomes between different ISGPS grades. B-POPF were divided into 2 subgroups as B1(without invasive procedures) and B2(with invasive procedures) then outcomes were analyzed across the subgroups. RESULTS: Biochemical leak(BL) did not differ from the non-fistula condition in all outcomes except postoperative hospital stay and cost. Non-fistula/BL,B-POPF and C-POPF condition differed significantly in terms of all clinical and economic outcomes except 30-day readmission rate.B1 differ from B2 subgroup greatly in terms of most critical terms of outcomes such as hemorrhage(15.2% vs. 34.3%,P=0.045),biliary fistula(13.0% vs 34.3%,P=0.023),postoperative hospital stay(32.0 d vs. 39.0 d,P=0.011). CONCLUSION: The present study has confirmed the effectiveness of the 2016 ISGPS definition and grading scheme in identifying three conditions that differ in terms of clinical and economic outcomes. Subclassification of B-POPF according to whether invasive procedures has been used has potential implications for accurate reporting and performance evaluation.