Renal resistive index by point of care ultrasound to predict sepsis associated acute kidney injury in critically ill children.

Abstract

Sepsis associated acute kidney injury (AKI) is linked with adverse outcomes in the PICU. Doppler-based renal resistive index (RRI) has shown promising results in adults for prediction of AKI. We aimed to explore the performance of RRI in children with sepsis. This prospective observational study (March - November 2022) included children aged 1-12years with sepsis admitted to the PICU. RRI and urine neutrophil gelatinase associated lipocalin (NGAL) were measured within 12h of admission. Children were followed up for 3days. AKI (new and persistent) was defined as any child with KDIGO stage 2 or 3 AKI on day 3. We enrolled 90 children but included 79 in final analysis. Two thirds (n = 53, 67%) had septic shock. Median (IQR) age was 6.2years (4.1-9.2). RRI decreased with increasing age. Twenty-six (33%) children had AKI on day 3. Mean (SD) RRI was higher in the AKI group [0.72 (0.08) vs. 0.65 (0.07), p < 0.001].The area under ROC curve for RRI to detect AKI among the 1-4year old group was 0.75 (95% CI:0.51, 0.98; p = 0.05) and among the 5-12year old group was 0.76 (0.62, 0.89; p = 0.001). An RRI 0.71 predicted AKI with 100% sensitivity and 46.2% specificity among the 1-4-year-old group and RRI 0.69 predicted it with 70% sensitivity and 77.5% specificity in the 5-12-year-old group. RRI and eGFR at admission were independent predictors of AKI on multivariable analysis. Urine NGAL 94.8ng/ml predicted AKI with 76.9% sensitivity and 77.4% specificity and AUROC was 0.74 (0.62, 0.86) among the 1-12-year-old group. RRI values varied with age. RRI showed good diagnostic accuracy to detect new/persistent AKI on day 3 in children with sepsis; however, it was less precise as an independent predictor.