Reappraisal of post-pancreatectomy hemorrhage (PPH) classifications: do we need to redefine grades A and B?

Background: Post-pancreatectomy hemorrhage (PPH) remains a major complication. The aim of this study was to reappraise the International Study Group of Pancreatic Surgery (ISGPS) classification. Methods: The clinical utility of the ISGPS classification was tested on consecutive pancreatic resections performed at the Pancreas Institute of the University of Verona Hospital. Results: PPH occurred in 65 of the 2429 patients (6.8%) undergoing pancreatic resection. Outcome of patients without PPH and with grade A PPH were comparable in terms of mortality, length of stay, ICU stay and readmission. Patients with grade B late and mild and grade B early and severe PPH had similar hospital stay and mortality rates, but differed in relaparotomy rate (10.1 vs. 81.2%, p < 0.01). Replacing " time of PPH onset" criterion with post-operative pancreatic fistula (POPF), severe PPH alone, mild PPH/POPF and severe PPH/POPF differed significantly for hospital stay (14 vs. 23 vs. 35 days, p < 0.01) and mortality rate (0 vs. 4 vs. 25%, p = 0.05). Conclusion: Grade A PPH shared the same outcome of patients without PPH. Grade B PPH included two categories of patients with different treatment modalities. The use of "concomitant POPF" instead of " time of onset" segregated three discrete categories that differed significantly in terms of clinical outcomes and management.