Predictors of Mortality in Pediatric Cardiac Intensive Care Units: A Systematic Review and Exploratory Meta-Analysis

Predictors of mortality in pediatric cardiac intensive care populations remain incompletely characterized. This study aimed to systematically evaluate and synthesize adjusted risk factors across relevant clinical conditions. A synthesis of existing evidence was done to identify the most reliable factors for risk stratification and to guide management in this vulnerable population. A systematic review and meta-analysis were conducted following PRISMA guidelines. A detailed search in PubMed, Embase, EBSCO, and OVID Medline was conducted for studies that were published from January 2020 through March 2026. The search yielded 866 citations. Studies reporting adjusted odds ratios (ORs) or hazard ratios (HRs) for predictors of adverse outcomes in pediatric cardiac intensive care settings were included. Random-effects meta-analysis was performed for predictors reported in at least two studies. Heterogeneity was assessed using the I2 statistic. A total of 14 studies, including 60923 patients, met the inclusion criteria. The pooled mortality rate across included studies was 8.98% (95% CI: 4.94-15.78%). A total of 30 predictors were identified. Notable findings, including pooled analyses, demonstrated that extracorporeal membrane oxygenation (ECMO) was associated with increased odds of adverse outcomes (pooled OR 2.26 (1.44-3.55)). Renal replacement therapy showed a strong association (pooled OR 6.7731 (4.3174-10.6255)), albeit with low heterogeneity. Vasoactive therapy and cardiac arrest were also significant predictors, with pooled ORs ranging from approximately 2.7 to 4. Unplanned reoperation demonstrated consistent effects with low heterogeneity, with a pooled OR of 2.3583 (1.4687-3.7867). Key clinical interventions and markers of disease severity, including ECMO, renal replacement therapy, vasoactive therapy, and cardiac arrest, are strongly associated with mortality in pediatric cardiac intensive care populations. These findings may inform risk stratification and clinical decision-making.