부산대학교 도서관

Background: Renal Angina Index (RAI) is a bedside tool for risk stratification of patients to predict acute kidney injury (AKI). Kidney biomarkers are better indicators of real-time injury and give us lead time for diagnosing impending AKI. Methods: We enrolled consecutive children aged 2 months–14 years admitted to a tertiary hospital in northern India over 2 years. RAI was calculated on day 0 (D0) and urinary (u) and plasma (p) neutrophil gelatinase-associated lipocalin (NGAL) were measured within 6 h of admission. Children were followed for the development of severe AKI on day 3 (D3) using Kidney Disease Improving Global Outcomes criteria to define and stage AKI. Results: Of the 253 children enrolled and analysed, 44 (17.4%) developed D3-AKI (stage 1 in 52.2%, stage 2 in 20.5% and stage 3 in 27.3%). Renal angina (RAI ≥ 8) on D0 was present in 66.7% children who developed stage 2/3 D3-AKI vs. 43.5% in children who did not develop D3-AKI /stage 1 AKI (p = 0.065). Area under ROC (AUROC) curve for D0-RAI to predict D3-severe-AKI was 0.66 (95% CI, 0.55–0.77). AUROC curve for uNGAL and pNGAL to predict D3-severe-AKI was 0.62 (95% CI, 0.50–0.74) and 0.48 (95% CI, 0.35–0.61), respectively. The severe AKI group had greater requirement of ventilation and inotropic support with mortality being thrice higher compared to the non-AKI group. Conclusion: RAI ≥ 8 and uNGAL had a high negative predictive value but low sensitivity for predicting D3-severe-AKI. pNGAL had a poor predictive value for D3-severe-AKI. A higher resolution version of the Graphical abstract is available as Supplementary information [ABSTRACT FROM AUTHOR]