The clinical utility window for acute kidney injury biomarkers in the critically ill

Table 3 Multivariable logistic regression for biomarker prediction of mortality or dialysis after addition to the reference model

Peak biomarkers within 24 h (log ₁₀)*	Beta coefficient	Odds ratio (95% CI)	P -value
Plasma cystatin C	2.56	12.9 (0.51 to 326)	0.12
Plasma NGAL	0.97	2.62 (0.32 to 21.5)	0.37
Urinary cystatin C/urinary creatinine	0.57	1.77 (0.85 to 3.72)	0.13
Urinary NGAL/urinary creatinine	1.05	2.87 (1.32 to 6.26)	0.008
Urinary alkaline phosphatase/urinary creatinine	0.52	1.68 (0.60 to 4.71)	0.32
Urinary GGT/urinary creatinine	0.46	1.59 (0.52 to 4.81)	0.41
Urinary α-GST/urinary creatinine	0.09	1.10 (0.72 to 1.66)	0.66
Urinary α-GST/urinary creatinine	0.63	1.89 (0.94 to 3.79)	0.07
Urinary albumin/urinary creatinine	1.00	2.72 (1.14 to 6.48)	0.02

*Each biomarker was added separately to the reference model. Urinary biomarkers were normalised to urinary creatinine. The reference model includes variables with P <0.1 on univariate analysis, namely the modified acute physiological and chronic health evaluation-II score (without creatinine score) and the presence of AKI₂₄. NGAL, neutrophil gelatinase-associated lipocalin; GGT, gamma-glutamyltranspeptidase; GST, glutathione-S-transferase.

ISSN: 1364-8535