Table 5.

Expected and observed numbers of major adverse kidney events at 1 month (MAKE30) and 1 year (MAKE365), stratified by deciles of logistic regression score (1 = lowest, 10 = highest), in the training and validation datasets.

MAKE30	Training sample			Validation sample
	Number of patients	Number of patients with MAKE30 expected	Number of patients with MAKE30 observed	Number of patients	Number of patients with MAKE30 expected	Number of patients with MAKE30 observed
1	94	10.9	11	40	5.3	8
2	90	16.9	13	39	8.7	11
3	91	22.8	25	40	11.4	11
4	91	29.4	34	39	13.6	13
5	92	36.2	34	39	16.2	15
6	91	42.6	41	40	18.7	12
7	92	50.9	52	39	20.5	21
8	91	58.3	61	40	24.5	26
9	92	67.2	62	39	27.3	26
10	91	76.8	79	39	31.9	35
Total	915	412	412	394	178.1	178
Hosmer-Lemeshow test statistic*			5.04, P=0.75			9.16, P = 0.52
AROC			0.77			0.74
MAKE365	Number of patients	Number of patients with MAKE365 expected	Number of patients with MAKE365 observed	Number of patients	Number of patients with MAKE365 expected	Number of patients with MAKE365 observed
1	91	18.2	18	39	10.9	13
2	91	31.9	34	39	16.1	13
3	91	41.3	42	39	19.5	18
4	91	49.4	48	39	22.0	21
5	90	55.5	56	39	24.0	27
6	91	62.0	59	39	25.8	26
7	91	67.6	65	39	27.5	27
8	91	73.1	76	39	29.3	32
9	91	77.9	82	39	31.5	30
10	90	83.1	80	38	33.5	33
Hosmer-Lemeshow test statistic*			4.73, P = 0.79			4.30, P= 0.93
C statistic			0.768 (95%CI 0.74–0.80)			0.717 (95%CI 0.67–0.77

^*Follows a X² distribution with 8 degrees of freedom for the training sample and 10 degrees of freedom for the validation sample.