Risk factors and predictive models for perioperative acute kidney injury in children: a narrative review

Background and Objective: Perioperative acute kidney injury (AKI) is a serious complication in children, with an incidence of 5–30% and up to 40% in neonates after cardiac surgery. It increases mortality and the risk of chronic kidney disease. This narrative review synthesizes current evidence on risk factors and predictive models for perioperative AKI in children, aiming to inform early risk stratification and preventive care.

Methods: A literature search was conducted up to March 2024 using PubMed/MEDLINE, Embase, Web of Science, and Cochrane Library. The search combined terms related to AKI, pediatrics, the perioperative period, risk factors, and prediction models. Studies focusing on pediatric patients (≤18 years) were included.

Key Content and Findings: Key risk factors include young age, congenital heart disease, exposure to nephrotoxic medications, and major surgeries like those using cardiopulmonary bypass. The review evaluates predictive models, from traditional statistical methods to machine learning models that incorporate novel biomarkers such as neutrophil gelatinase-associated lipocalin and kidney injury molecule-1 for earlier detection. Promising biomarkers like urinary L-FABP and TIMP-2×IGFBP7 are also highlighted. Integrating these tools into clinical workflows can guide proactive management.

Conclusions: Early identification of high-risk children is crucial. While predictive modeling is advancing, a gap remains in models specifically validated for pediatric populations. Future research should focus on multicenter studies to refine pediatric-specific models, validate novel biomarkers, and develop personalized approaches. Implementing evidence-based, risk-stratified care has the potential to significantly improve outcomes and long-term renal health for children undergoing surgery.