Diagnosis and outcomes of acute kidney injury using surrogate and imputation methods for missing preadmission creatinine values

Table 2 Acute kidney injury agreement between surrogate methods for missing baseline creatinine and preadmission serum creatinine

	AKI (%) n = 498	Sensitivity	Specificity	Kappa	McNemar*
Preadmission creatinine	25.1 (95% CI 21.5-29.1%; n = 125)	-	-	-
First serum creatinine at hospital admission	22.5 (95% CI 19.0-26.4%; n = 112)	87.2	99.2	0.895 (95% CI 0.848-0.942)	0.004
Minimal creatinine within 2 weeks after ICU admission	43.2 (95% CI 38.9-47.6%; n = 215)	98.4	75.3	0.595 (95% CI 0.526-0.664)	<0.0001
MDRD computed serum creatinine	26.7 (95% CI 23.0-30.8%; n = 133)	92.8	95.4	0.864 (95% CI 0.813-0.915)	0.169
CKD-EPI computed serum creatinine	27.1 (95% CI 23.4-31.2; n = 135)	93.6	95.1	0.865 (95% CI 0.814-0.916)	0.076

*p < 0.01 represented significant difference, with the Bonferroni correction
Misclassification rates were 3.8% for the first SCr (overestimation 0.6% and underestimation 3.2%), 18.9% for the minimal SCr (overestimation 18.5% and underestimation 0.4%), 5.2% with MDRD (overestimation 3.4% and underestimation 1.8%), 5.2% with CKD-EPI (overestimation 3.6% and underestimation 1.6%)
MDRD: Modification of Diet in Renal Disease
CKD-EPI: Chronic Kidney Disease Epidemiology Collaboration

ISSN: 1471-2369