Abstract Background There is limited data on applying urinary biomarkers for prediction of kidney outcomes in pediatric acute kidney injury (AKI). Methods We prospectively measured urinary neutrophil gelatinase-associated lipocalin (NGAL), tissue metalloproteinases-2 (TIMP-2), insulin-like growth factor-binding protein 7 (IGFBP-7) and C–C motif chemokine ligand 14 (CCL14), alongside serum kidney function test in critically ill children with AKI admitted to the pediatric intensive care unit. The primary outcomes included persistent AKI (lasting for ≥ 72 h) and prolonged AKI (lasting for ≥ 7 days). Results There were altogether 134 patients (median age 4.3 years; 43.3% female; AKI severity stage 1: 44.8%, stage 2: 33.6% and stage 3: 21.6%). The incidence of persistent and prolonged AKI was 40.3% and 25.4%, respectively. All four biomarkers, either measured singly, simultaneously or serially, significantly predicted both outcomes, with NGAL demonstrating the best performance (areas under the curve AUC 0.72 0.61, 0.83 for persistent AKI and 0.72 0.61, 0.84 for prolonged AKI). Integrating the simultaneous AKI staging with biomarker levels significantly improved prediction (NGAL: AUC 0.86 0.78, 0.94 for persistent AKI and 0.87 0.79, 0.96 for prolonged AKI). Persistent AKI increased the risk of acute kidney disease (hazard ratios HR: 2.59 1.55, 4.34), which was associated with kidney function non-recovery 90 days after AKI (HR 7.73 1.01, 59.03). Conclusions Urinary NGAL, TIMP-2, IGFBP-7 and CCL14 demonstrated promising performance of predicting kidney function non-recovery within 7 days of AKI onset. Integrating urinary biomarkers with concurrent clinical data substantially enhanced predictive performance. Graphical Abstract
Hui et al. (Wed,) studied this question.