OBJECTIVES: It is widely accepted that postoperative pancreatic fistula (POPF) accompanied by bacterial infection results in a worse outcome than POPF alone. However, few studies evaluating predictive indicators of POPF have focused on bacterial infection. METHODS: A consecutive 100 patients who underwent pancreaticoduodenectomy at our institute for periampullary disease were enrolled. POPF was assessed according to the International Study Group of Pancreatic Fistula consensus guidelines; grades B and C were defined as clinically relevant POPF (CR-POPF). The patients' characteristics, perioperative surgical factors, and laboratory data including the results of culture and smear testing performed using drainage fluid on postoperative days (PODs) 1 and 3 were analyzed. RESULTS: The overall incidence of CR-POPF was 25%. Univariate analyses revealed that the factors associated with CR-POPF were male sex, soft pancreas, MPD diameter, higher serum C-reactive protein concentration and white blood cell count on POD 3, higher amylase concentration in drainage fluid, and culture and/or smear positivity of drainage fluid. Multivariate analysis newly revealed that the smear positivity of drainage fluid on POD 3 was the independent risk factors for CR-POPF (p = 0.027). CONCLUSIONS: Smear positivity of drainage fluid on POD 3 after pancreaticoduodenectomy may be a new predictor of CR-POPF.
OBJECTIVES: It is widely accepted that postoperative pancreatic fistula (POPF) accompanied by bacterial infection results in a worse outcome than POPF alone. However, few studies evaluating predictive indicators of POPF have focused on bacterial infection. METHODS: A consecutive 100 patients who underwent pancreaticoduodenectomy at our institute for periampullary disease were enrolled. POPF was assessed according to the International Study Group of Pancreatic Fistula consensus guidelines; grades B and C were defined as clinically relevant POPF (CR-POPF). The patients' characteristics, perioperative surgical factors, and laboratory data including the results of culture and smear testing performed using drainage fluid on postoperative days (PODs) 1 and 3 were analyzed. RESULTS: The overall incidence of CR-POPF was 25%. Univariate analyses revealed that the factors associated with CR-POPF were male sex, soft pancreas, MPD diameter, higher serum C-reactive protein concentration and white blood cell count on POD 3, higher amylase concentration in drainage fluid, and culture and/or smear positivity of drainage fluid. Multivariate analysis newly revealed that the smear positivity of drainage fluid on POD 3 was the independent risk factors for CR-POPF (p = 0.027). CONCLUSIONS: Smear positivity of drainage fluid on POD 3 after pancreaticoduodenectomy may be a new predictor of CR-POPF.