Skip to main content
. 2016 Mar 3;11(3):e0149706. doi: 10.1371/journal.pone.0149706

Table 6. Logistic regression model for clotting during intermittent hemodialysis.

Model A: simple model; Model B: mixed effect for repeated measures model.

Variable P OR (95% CI)
Model A (simple)*
Heparin (yes/no) 0.047 0.57 (0.33–0.98)
Malfunctioning catheter < 0.0001 3.59 (2.24–5.77)
System pressure increase over 25% < 0.0001 2.59 (1.61–4.17)
Model B (mixed effect for repeated measures model)* *
Platelets < 100,000/mm3 0.047 0.28 (0.08–0.98)
Heparin (yes/no) 0.032 0.48 (0.25–0.94)
Malfunctioning catheter < 0.0001 2.92 (1.77–4.80)
System pressure increase over 25% 0.001 2.17 (1.40–3.38)

* Model performance: area under ROC curve = 0.71 (0.65–0.77); Hosmer-Lemeshow χ2 = 0.76.

** Model performance: area under ROC curve = 0.72 (0.66–0.77); ROC, receiver operating characteristic; OR, odds ratio; CI, confidence interval.