Preoperative plasma growth-differentiation factor-15 for prediction of acute kidney injury in patients undergoing cardiac surgery

Table 3 Reclassification analysis

Outcome: absent
Updated model (risk categories)
Initial model (risk categories)	(0, 0.01)	(0.01, 0.05)	(0.05, 0.1)	(0.1, 1)	% reclassified
(0, 0.01)	0	0	0	0	-
(0.01, 0.05)	0	53	2	0	4
(0.05, 0.1)	0	12	143	13	15
(0.1, 1) 0	0	24		669	3
Outcome: present
Updated model (risk categories)
Initial model (risk categories)	(0, 0.01)	(0.01, 0.05)	(0.05, 0.1)	(0.1, 1)	% reclassified
(0, 0.01)	0	0	0	0	-
(0.01, 0.05)	0	4	1	0	20
(0.05, 0.1)	0	1	16	1	11
(0.1, 1)	0	0	1	233	0
Combined data
Updated model (risk categories)
Initial model (risk categories)	(0, 0.01)	(0.01, 0.05)	(0.05, 0.1)	(0.1, 1)	% reclassified
(0, 0.01)	0	0	0	0	-
(0.01, 0.05)	0	57	3	0	5
(0.05, 0.1)	0	13	159	14	15
(0.1, 1)	0	0	25	902	3

Net-reclassification improvement (NRI) (categorical) (95 % CI): 0.0229 (0.0014, 0.0445); p value 0.03697. NRI (continuous) (95 % CI): 0.308 (0.1739, 0.4421); p value: 0.00001. Integrated discrimination improvement (95 % CI): 0.015 (0.006, 0.024); p value: 0.00107. Reclassification table of the model without (initial) and with (updated) growth-differentiation factor-15 (GDF-15) as a predictor of cardiac-surgery-associated acute kidney injury (CSA-AKI). The addition of GDF-15 in the logistic regression model significantly improved prediction of the development of CSA-AKI.

ISSN: 1364-8535