Table 3.

Predictive capacity of the three multiple binomial logistic regression analysis to predict 30-day mortality

	First model	Second model	Third model
Overall predictive capacity	85.4 %	82.3 %	76.0 %
Area under curve; p-value	0.79; p < 0.001	0.74; p < 0.001	0.59; p = 0.03
Sensitivity (95 % CI)	64 (43–82)	56 (35–76)	22 (9–42)
Specificity (95 % CI)	93 (84–98)	92 (83–97)	96 (89–99)
Positive likelihood ratio (95 % CI)	9.1 (3.7–22.2)	6.6 (2.9–15.4)	5.4 (1.5–20.1)
Negative likelihood ratio (95 % CI)	0.4 (0.2–0.7)	0.5 (0.3–0.8)	0.8 (0.7–1.0)
Positive predicted value (95 % CI)	76 (53–92)	70 (46–88)	67 (30–93)
Negative predicted value (95 % CI)	88 (78–94)	86 (76–93)	77 (67–85)

CI confidence interval