Background: Postoperative pancreatic fistula (POPF) remains a leading cause of morbidity after pancreaticoduodenectomy (PD). In the present study we sought to establish a preoperative scoring system with which to predict this complication.
Patients and methods: The clinical records of 387 consecutive patients who underwent PD for periampullary tumor between 2004 and 2009 were reviewed retrospectively. Patients were divided into two groups; 279 consecutive patients constituted the study group and the next 108 patients constituted the validation group. Univariate and multivariate logistic regression analyses were performed using preoperative and surgical factors potentially influencing grade B or C POPF in the study group, and a score to predict POPF was constructed. This score was confirmed in the validation group.
Results: In the study group, grade A POPF was recognized in 45 patients (16%), grade B in 98 (35%), and grade C in 5 (2%). A preoperative predictive scoring system for POPF (0-7 points) was constructed using the following 5 factors; main pancreatic duct index <0.25 (2 points), away from portal vein on computed tomography (2 points), disease other than pancreatic cancer (1 point), male (1 point), and intra-abdominal thickness >65 mm (1 point). The nomogram showed an area under the curve (AUC) of 0.808. This scoring system was highly predictive for grade B or C POPF in the validation group (AUC = 0.834).
Conclusions: The present scoring system satisfactorily predicted the occurrence of POPF and thus will be useful for the perioperative risk management of patients undergoing PD in a high-volume center hospital.