Abstract

Acute kidney injury (AKI) is a significant clinical concern in neonates, threatening optimal outcomes. Early and accurate diagnosis is crucial; however, current methods lack sufficient sensitivity. This meta-analysis aimed to evaluate urinary kidney injury molecule-1 (uKIM-1) for AKI in neonates by quantifying differences in uKIM-1 levels between AKI and non-AKI neonates. We systematically searched major databases for comparative studies. Quality assessment was performed using the Newcastle-Ottawa Scale, and the certainty of the evidence was assessed according to the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology. A random-effects meta-analysis estimated the pooled Hedges’ g in uKIM-1 levels, accounting for heterogeneity. Subgroup analyses explored sources of heterogeneity (continent, study design, sampling time, AKI definition). Publication bias was assessed using Egger’s and Begg’s tests, as well as with a funnel plot. Data from 13 studies involving 552 neonates indicated a significant association between elevated uKIM-1 levels and AKI. High heterogeneity was observed (I² = 80.32%). The pooled Hedges’ g was 0.62 (95% CI: 0.16–1.07, p = 0.01). Subgroup analysis showed stronger associations in African studies (Hedges’ g = 2.12), those using KDIGO (Hedges’ g = 0.96), cohort studies, and sampling within 2–4 days (Hedges’ g = 0.76). No publication bias was detected. This meta-analysis synthesizes evidence on uKIM-1 as an AKI biomarker. While uKIM-1 shows promise, high heterogeneity and diagnostic performance warrant further research to improve AKI detection and management in neonates.