Pancreas-visceral fat CT value ratio and serrated pancreatic contour are strong predictors of postoperative pancreatic fistula after pancreaticojejunostomy

Table 3 Univariate and multivariate analysis for evaluating intraoperative risk factors associated with POPF

Variables	non-POPF, BL (n = 235)	POPF Grade BC (n = 27)	P-value (Univariate)	Odd’s ratio (95% CI)	P-value (Multivariate)
Procedure: PPPD/SSPPD/PD	6/203/26	0/25/2	0.577
Laparoscopic surgery (yes/no)	9/226	3/24	0.087
Combined PV resection (yes/no)	107/128	4/23	0.002	2.72 (0.80–9.31)	0.110
Combined artery resection (yes/no)	9/226	3/24	0.087
Combined distal pancreatectomy (yes/no)	3/232	1/26	0.330
Pancreatic texture judged intraoperatively (soft/hard)	102/133	21/6	0.001	2.89 (1.00–8.35)	0.050
Diameter of main pancreatic duct judged intraoperatively	4 (1–15)	3 (2–8)	0.054
Pancreatic stent (yes/no)	123/112	17/10	0.654
Operation time (min)	498.5 (286-1373^a)	496.0 (333–670)	0.591
Intraoperative blood loss (ml)	713.0 (20-20983^b)	692.0 (210–5522)	0.234

^aThis case underwent SSPPD, transverse colectomy and low anterior resection for triple cancer (duodenal papilla, transverse colon and rectum)
^bThis case developed intraoperative massive bleeding due to the presence of intraabdominal abscess and severe adhesion to adjacent organs and vessels, but finally recovered
CI confidence interval, PPPD pylorus-preserving pancreaticoduodenectomy, SSPPD subtotal stomach-preserving pancreaticoduodenectomy, PD pancreaticoduodenectomy, PV portal vein, POPF postoperative pancreatic fistula, BL biochemical leak, Statistical analysis: Mann- Whitney U test for contentious variables. χ2 analysis for discrete variables

ISSN: 1471-2482