Skip to main content
. 2022 Jun 28;9:850929. doi: 10.3389/fnut.2022.850929

TABLE 2.

Univariate and multivariate analysis of postoperative complications.

Factors Training group
Validation group
Univariate analysis
Multi- analysis
c value
Univariate analysis
Multi- analysis
P-value
OR (95%CI) P-value OR (95%CI) P-value
Radiomic sarcopenia 3.3 (1.8, 6.0) <0.001 0.003* 3.6 (2.0, 6.5) <0.001 <0.001*
Gender 0.197 0.767
Female Ref Ref
Male 0.7 (0.3, 1.2) 0.197 0.9 (0.5, 1.7) 0.767
Age > 65 years 2.3 (1.3, 4.1) 0.006 0.142 1.9 (1.0, 3.4) 0.037 0.135
BMI < 18.5 kg/m2 1.8 (0.7, 4.5) 0.198 2.7 (1.1, 6.6) 0.028 0.622
NRS-2002 score 0.106 <0.001 0.013*
1–2 Ref Ref
3–4 1.3 (0.7, 2.5) 0.39 2.6 (1.4, 4.9) 0.026 0.005
5 2.8 (1.0, 7.5) 0.04 4.8 (1.5, 15.2) 0.079 0.112
Laparoscopic surgery 0.8 (0.4, 1.4) 0.018 0.032* 0.8 (0.4, 1.4) 0.397
Total gastric resection 1.6 (0.9, 2.8) 0.135 1.8 (1.0, 3.2) 0.040 0.762
Combined resection 2.7 (1.2, 6.2) 0.015 0.124 6.1 (2.5, 14.6) < 0.001 <0.001*
Anastomotic type 0.007 0.177 0.013 0.028*
Roux-en-Y Ref Ref
Billroth I 0.5 (0.2, 0.9) 0.025 0.154 0.6 (0.3, 1.1) 0.093 0.316
Billroth II 1.7 (0.8, 3.8) 0.162 0.450 0.3 (0.1, 0.7) 0.008 0.010
TNM stage 0.173 0.007 0.050*
I Ref Ref
II 1.5 (0.6, 3.5) 0.393 3.3 (1.5, 7.5) 0.004 0.015
III 1.9 (1.0, 3.8) 0.065 2.5 (1.2, 5.1) 0.011 0.182

BMI, body mass index; NRS-2002, nutritional risk screening; TNM, tumor–node–metastasis; OR, odds ratio; CI, confidence interval; ref, reference.

*Statistically significant in multivariate analysis.