Skip to main content
. 2016 Jul 28;6:30271. doi: 10.1038/srep30271

Table 4. Evolution of odds ratio [95% Wald Confidence Limits] of eGFR with the outcome stroke/TIA or death at 1 year.

Model eGFR CKD-EPI <30 vs ≥80 ml/min/1.73 m2 eGFR CKD-EPI 30–49 vs ≥80 ml/min/1.73 m2 eGFR CKD-EPI 50–79 vs ≥80 ml/min/1.73 m2 Overall Pr > ChiSq
Model 1 (a) 7.955 [4.104; 15.418] p < 0.0001 4.622 [2.680; 7.972] p < 0.0001 1.883 [1.142; 3.105] p = 0.0132 p < 0.0001
Model 2 (b) 6.834 [3.248; 14.377] p < 0.0001 4.185 [2.260; 7.749] p < 0.0001 1.912 [1.083; 3.376] p = 0.0255 p < 0.0001
Model 3(c) 5.616 [2.258; 13.971] P = 0.0002 4.135 [1.902; 8.991] P = 0.0003 2.247 [1.117; 4.520] p = 0.0231 P = 0.0004
Model 4 (d) 4.699 [2.113; 10.449] P = 0.0001 3.004 [1.551; 5.819] P = 0.0011 1.792 [0.978; 3.285] p = 0.0591 P = 0.0003

(a) Model 1 includes age and sex. Hosmer and Lemeshow Goodness-of-Fit p-value = 0.2248. (b) Model 2 includes age, sex and co-morbidities: coronary artery disease, chronic heart failure, previous stroke, previous TIA, ischaemic thrombo-embolic complications, haemorrhagic events and malignancy. Hosmer and Lemeshow Goodness-of-Fit p-value = 0.1183. (c) Model 3 is the model 2 with confounding factors significant (p < 0.1) in the univariate model, i.e. nation, setting, AF type, valvular heart disease, cardiomyopathy, other cardiac disease, COPD, peripheral vascular disease, diabetes, alcohol >= 2–3/day, physical activity, previous pharmacological cardioversion, electrical cardioversion, catheter ablation, EHRA symptoms - BMI - SBP - DBP - Management Strategy - Main reason for admission/consultation. Hosmer and Lemeshow Goodness-of-Fit p-value = 0.7486. (d) Model 4 is the model 2 with confounding factors significant (p < 0.1) in the univariate model and kept into the model with the stepwise procedure, i.e. nation, diabetes, physical activity and main reason for admission/consultation. Hosmer and Lemeshow Goodness-of-Fit p-value = 0.6485.