Table 3.

Univariate and multivariate analysis of variables found to be significantly associated with morbidity

	Patients (n = 652)	No. of complications	Univariate unadjusted odds ratio (95% CI)	Multivariate adjusted odds ratio (95% CI)
Age (per 10-year increment)			1.18 (1.02–1.37)
Sex
Female	293	132 (45.1)	1.00
Male	359	199 (55.4)	1.51 (1.06–2.15)
BMI (per point increment)			1.05 (1.01–1.10)
Hypertension
No	628	305 (48.6)	1.00
Yes	24	15 (63)	1.63 (1.04–2.55)
Cardiac history
No	616	296 (48.1)	1.00
Yes	36	22 (61)	1.74 (1.13–2.70)
Pulmonary history
No	552	267 (48.4)	1.00
Yes	100	68 (68)	2.29 (1.32–3.98)	2.05 (1.15–3.67)
Blood loss (per 100 ml increment)			1.02 (1.00–1.04)
Type of tumor
Pancreatic adenocarcinoma	233	107 (46)	1.00
Ampulla of Vater adenocarcinoma	157	93 (59)	1.72 (1.08–2.73)	1.73 (1.07–2.80)
Distal bile duct adenocarcarcinoma	91	41 (45)	0.96 (0.55–1.68)
Pancreatitis	72	34 (47)	1.05 (0.57–1.94)
Duodenum adenocarcinoma	14	9 (64)	2.36 (0.69–8.10)
Other malignancy	46	22 (48)	1.06 (0.53–2.14)
Other benign	26	15 (58)	1.53 (0.64–3.67)
Tubulovillus adenoma	13	10 (77)	4.71 (0.97–22.81)

Numbers in parentheses are percentages unless indicated otherwise

Factors analyzed in univariate analysis that were not significant include: diabetes, American Society of Anesthesiologists classification, surgeon’s experience, classical Whipple or pylorus-preserving pancreatoduodenectomy, transsection with surgical knife or linear stapler, use of multicomponent fibrin sealant, single or Roux-en-Y jejunal loop, one- or two layer anastomosis, drainage of pancreatic or biliary duct, packed cells transfused, use of octreotide and microscopic completeness of pancreatic resection plane in case of malignancy. Pancreatic texture was not scored regularly and was therefore omitted from the analysis