Prognostic performance of immune-metabolic indices compared with severity scores in children requiring continuous kidney replacement therapy for acute kidney injury.

BACKGROUND

Acute kidney injury (AKI) requiring continuous kidney replacement therapy (CKRT) is associated with substantial mortality in critically ill children. Although illness severity scores provide robust prognostic information, the prognostic value of readily available immune-metabolic indices in this population remains uncertain. We aimed to evaluate the prognostic performance of immune-metabolic indices measured at CKRT initiation and compare their performance with established severity scores.

METHODS

This retrospective single-center cohort study included children younger than 18 years who underwent CKRT for KDIGO-defined AKI between January 2020 and December 2024. Lactate-to-albumin ratio (LAR), urea-to-creatinine ratio (UCR), neutrophil-to-lymphocyte ratio (NLR), and monocyte-to-lymphocyte ratio (MLR) were calculated using laboratory values obtained within 24 h before CKRT initiation. The primary outcome was 28-day mortality following CKRT initiation. Receiver operating characteristic (ROC) analysis and multivariable logistic regression were used to evaluate prognostic performance.

RESULTS

Ninety patients were included, and 19 (21.1%) died within 28 days after CKRT initiation. Among the evaluated immune-metabolic indices, LAR demonstrated the strongest discriminative performance for 28-day mortality (AUC 0.779, 95% CI 0.659-0.898), closely approximating PELOD-2 (AUC 0.785) and PRISM III (AUC 0.790) scores. UCR and MLR also demonstrated significant discriminatory ability (AUC 0.721 and 0.719, respectively). In univariable analyses, LAR and UCR were significantly associated with mortality. However, neither biomarker remained statistically significant after adjustment for PELOD-2 score.

CONCLUSIONS

LAR and UCR were associated with 28-day mortality, whereas MLR showed a borderline inverse association. Although these biomarkers demonstrated discriminative performance approaching that of established severity scores, they did not provide independent prognostic information beyond PELOD-2 score. Immune-metabolic indices may therefore serve as complementary bedside markers of physiologic burden rather than substitutes for established severity-of-illness scores.