Clinical Profile and Survival Outcomes of Children with Acute Kidney Injury Requiring Renal Replacement Therapy: A Retrospective Study from a Tertiary Pediatric Intensive Care Unit

Background and aims: Acute kidney injury (AKI) in children is a serious and potentially life-threatening condition, particularly among critically ill patients.Despite advances in renal replacement therapy (RRT), mortality remains high.This study aimed to evaluate the etiology and indications for RRT in pediatric AKI, identify predictors of mortality, and compare survival outcomes between hemodialysis (HD) and peritoneal dialysis (PD).Patients and methods: This single-center, retrospective observational study included children aged 1 month to 18 years who received RRT for AKI at a tertiary care hospital in Eastern India between January 2023 and December 2024.Acute kidney injury was defined by the Kidney Disease: Improving Global Outcomes (KDIGO) 2012 criteria.Demographic, clinical (including illness severity scores), and laboratory data were extracted from medical records.Predictors of mortality were evaluated using univariate and multivariate logistic regression analysis.Kaplan-Meier estimation, the log-rank test, and Cox proportional hazards regression analysis were used to compare survival between HD and PD.Results: Out of 40 patients, 28 (70%) received HD and 12 (30%) underwent PD.The overall in-hospital mortality rate was 27.5% (11/40).The most common cause of AKI necessitating RRT was sepsis (39.5%), which included urosepsis and systemic sepsis.Refractory fluid overload (>10%) was the primary indication for RRT initiation in 67.5% of patients.Non-survivors had a significantly higher prevalence of sepsis, multiple organ dysfunction syndrome (MODS), underlying comorbidities, vasopressor use, and greater mean percentage of fluid overload (mean: 16.8% in non-survivors vs 8.3% in survivors; p = 0.003).In univariate analysis, dialysis modality, comorbidities, fluid overload, sepsis, MODS, and vasopressor use were significantly associated with mortality.However, none of these were independently predictive in the multivariate logistic regression analysis.Median time to in-hospital death or discharge was significantly longer in the HD group (88 vs 22 days, p = 0.0012).Cox regression showed an increased association with in-hospital mortality risk in the PD group hazard ratio (HR) = 4.41, 95% confidence interval (CI): 1.19-16.33,p = 0.026.Conclusion: Higher mortality was observed in patients with sepsis, MODS, and significant fluid overload.The higher mortality in the PD group is likely attributable to selection bias, as PD was preferentially used in sicker, hemodynamically unstable patients.