Skip to main content
. 2019 Feb 22;14(2):e0212750. doi: 10.1371/journal.pone.0212750

Table 4. Logistic regression analysis of kidney-transplant patients for rejection.

Predictor β SE β Wald x2 p eβ (OR) 95% CI for eβ
Lower Upper
HLA-G +3010 (Categorical variable) --- --- 11.054 0.004 --- --- ---
HLA-G +3010 CG 1.664 0.630 6.969 0.008 5.278 1.535 18.148
HLA-G +3010 GG -0.470 0.806 0.340 0.560 0.625 0.129 3.035
Constant -0.916 0.483 3.598 0.058 0.400
Overall model evaluation Null model Predicted model
-2 Log likelihood 87.720 75.304
Wald test 0.995 3.598
Coefficient constant 0.251 -0.916
Goodness-of-fit test p
Hosmer & Lemeshow 1.000
Cox and Snell R 0.176
Nagelkerke R2 0.236

Binary Logistic Regression (method: forward stepwise conditional). Analysis performed with genotypes for KTN (coded as “0”; n = 36) versus KTR (coded as “1”; n = 28). Predicted logit (equation) of rejection = (-0.916) + (1.664)* (HLA-G +3010 CG) + (-0.470) * (HLA-G +3010 GG). The contrast of categorical variables and their references were respectively the indicator and the last subcategory according to codes described in available file “Data set.xlsx”. HLA-G +3010 categorical variable had as reference HLA-G +3010 CC genotype. KTN: Kidney-transplant patients with no rejection. KTR: Kidney-transplant patients who developed episodes of rejection. OR: Odds Ratio. CI: Confidence Interval.