Renal Angina Indices and Urinary Biomarkers: A Combined Approach to Predict Acute Kidney Injury in Critically Ill Patients.

KEY POINTS: The modified renal angina index outperformed prior indices for early severe AKI prediction in critically ill adults. Combining modified renal angina index with [tissue inhibitor of metalloproteinase-2]×[IGF binding protein 7] improved area under the receiver operating characteristic curve, although not statistically significant, requiring external validation. Early AKI detection supports timely nephroprotective actions and individualized management in critical care. BACKGROUND: Several renal angina indices (RAIs) have been developed to stratify AKI risk in critically ill populations. However, limited evidence supports whether incorporating urinary biomarkers enhances predictive performance. This study aimed to evaluate the diagnostic accuracy of three RAIs, with and without urinary biomarkers integration, for predicting severe AKI (defined as Kidney Disease Improving Global Outcomes stages 2–3) in critically ill adults. METHODS: In this prospective, multicenter diagnostic study, we assessed the ability of three RAIs to predict severe AKI, defined as Kidney Disease Improving Global Outcomes stages 2–3, within 72 hours of intensive care unit admission. We further evaluated whether the addition of urinary biomarkers improve predictive accuracy, using area under the receiver operating characteristic curve (AUC) and net reclassification improvement (NRI). RESULTS: We enrolled 134 critically ill patients, of whom 15% developed severe AKI within 72 hours. Diagnostic performance of RAIs and urinary biomarkers alone was modest (AUC: Matsuura RAI, 0.63; Del Toro RAI, 0.71; and modified RAI [mRAI], 0.76). Incorporating urinary biomarkers improves the predictive performance of all RAIs, particularly enhancing specificity. Among biomarkers assessed, (tissue inhibitor of metalloproteinase-2 [TIMP‐2])×(IGF binding protein 7 [IGFBP7]) demonstrated the greatest capacity to reclassify patients toward event prediction across RAI models. The combination of mRAI with [TIMP‐2]×[IGFBP7] yields the highest absolute predictive performance (AUC, 0.82; 95% confidence interval, 0.72 to 0.93). CONCLUSIONS: Sequential risk stratification combining clinical RAIs with targeted urinary biomarkers profiling significantly improves early prediction of severe AKI in critically ill adults' patients. Although RAIs and urinary biomarkers alone show limited predictive accuracy, their integration—particularly mRAI plus [TIMP‐2]×[IGFBP7]—optimizes discrimination and may facilitate earlier intervention in high-risk intensive care unit adult patients.