Skip to main content

Table 4 Logistic regression model for CR-POPF occurrence in 233 unmatched patients

From: Modified duct-to-mucosa versus conventional pancreaticoenterostomy for pancreaticoduodenectomy: a retrospective cohort study based on propensity score matching analysis

Variable Univariate Multivariate
With CR-POPF, n (%) Without CR-POPF, n (%) P value Relative risk (95% CI) P value
PE approach    < 0.001   0.009
 End-to-side 17 (15.6) 92 (84.4)   1  
 Duct-to-mucosa 1 (0.8) 123 (99.2)   0.11 (0.02–0.57)  
BMI (kg/m2)    < 0.001   < 0.001
 < 24 5 (3.0) 161 (97.0)   Ref.
 24–28 6 (10.9) 49 (89.1)   11.89 (2.14–65.94) 0.005
 > 28 7 (58.3) 5 (41.7)   49.39 (5.20–469.19) < 0.001
Diabetes    < 0.001   0.001
 No 6 (3.2) 181 (96.8)   1  
 Yes 12 (26.1) 34 (73.9)   10.54 (2.55–43.61)  
Final diagnosis    0.016   0.324
 Pancreatic cancer 6 (4.3) 134 (95.7)   1  
 Other pathologic findings 12 (12.9) 81 (87.1)   2.24 (0.45–11.10)  
Pancreas texture    0.018   0.480
 Firm 8 (5.0) 153 (95.0)   1  
 Soft 10 (13.9) 62 (86.1)   1.81 (0.35–9.45)  
Pancreatic duct diameter    0.029   0.021
 3 mm 10 (5.4) 174 (94.6)   1  
 < 3 mm 8 (16.3) 41 (83.7)   6.72 (1.33–33.88)  
Neoplastic    0.175   0.865
 No 2 (20.0) 8 (80.0)   1  
 Yes 16 (7.2) 207 (92.8)   0.73 (0.02–26.69)