Le Sepsis pédiatrique en contexte à ressources limitées : profil épidémiologique, évolution et facteurs prédictifs de défaillance viscérale

Résumé

Introduction: Pediatric sepsis is a major health emergency and a leading cause of mortality worldwide. Severity scores provide improved tools for the identification of severe cases and prognostic assessment. Data are lacking in resource-limited settings. Aims: To determine the prevalence of pediatric sepsis among hospitalized children, describe its epidemio- logical and clinical features, and identify factors associated with multiple organ dysfunction (MODS). Materials and methods: This retrospective single-center study was conducted in the Department of Pedia- trics at Fattouma Bourguiba Hospital, from January 2021 to June 2025. Children aged 1 month to 14 years with sepsis, defined according to the Phoenix criteria, were included. Disease severity was assessed using PELOD-2. Epidemiological, clinical, biological, therapeutic, and outcome data were evaluated. Results: Up to 109 sepsis cases were identified (prevalence 1.04%), representing 12.9% of pediatric intensive care unit (PICU) admissions. Forty-eight patients met the inclusion criteria, with a mean age of 21.7 ± 38 months. Direct PICU admission occurred in 52.1%. Multiple organ dysfunction occurred in 70.8%, and mor- tality reached 41.7%. In univariate analysis, factors associated with MODS at PICU admission were failure to thrive, hypotension, oliguria, dysthermia, hypoxemia, hyperleukocytosis and elevated lactate. However, no independent predictors were identified in multivariate analysis. Indicators of disease severity, including the need for repeated fluid boluses and vasoactive drugs, were more frequently observed in patients with MODS. ROC curve analysis identified PELOD-2 ≥ 8 and Phoenix score ≥ 4 as thresholds more frequently observed in patients with MODS. Conclusion : Pediatric sepsis remains associated with high mortality and frequent organ dysfunction. PELOD-2 and Phoenix scores may be useful for early risk stratification. However, these findings should be considered exploratory and require external validation in larger studies, particularly in resource-limited settings.