RESUMEN
AIM: The performance of plasma neutrophil gelatinase-associated lipocalin (pNGAL) for prediction of acute kidney injury (AKI) in non-cardiac surgical patients has not been well described. This study investigates the use of pNGAL for early detection of AKI in patients admitted to an intensive care unit (ICU) after major or ultra-major non-cardiac surgery. METHODS: A total of 151 patients were recruited. Blood samples at 0 h and 6 h post-ICU admission were collected. Primary outcome was occurrence of AKI within 48 h of ICU admission defined using Acute Kidney Injury Network (AKIN) classification. RESULTS: Forty-five (29.8%) patients developed AKI within 48 h of ICU admission. Among them, 22, 14, and nine were classified as AKIN Stage 1, 2, and 3 respectively. pNGAL levels at 0 h and 6 h were significantly related to AKI severity. The AUROC for pNGAL at 0 h and 6 h increased with AKI severity (AKIN stage ≥1 0.671 ± 0.048 and 0.691 ± 0.047; stage ≥2 0.737 ± 0.055 and 0.796 ± 0.048; stage 3 0.829 ± 0.072 and 0.860 ± 0.065, respectively) and requirement of renal replacement therapy (0.880 ± 0.059 & 0.837 ± 0.088). Change of pNGAL from 0 h to 6 h showed no advantage in predictive power compared with pNGAL level at 0 h or 6 h alone. The addition of pNGAL into clinical AKI prediction model could only provide marginal benefit. CONCLUSION: pNGAL correlated with severity of AKI and requirement of renal replacement therapy in ICU patients who received major or ultra-major non-cardiac surgery. However, the benefit of adding pNGAL into clinical AKI prediction model is marginal.