Development and Validation of a Prediction Model for Early ARDS in Children with Sepsis in the PICU: A Multicenter Retrospective Cohort Study.

OBJECTIVE: To identify early factors associated with pediatric acute respiratory distress syndrome (PARDS) within 48 hours after admission in children with sepsis and to develop and validate a combined prediction model. METHODS: This retrospective cohort study included 547 children with sepsis admitted to the pediatric intensive care unit between June 2023 and December 2025 as the derivation cohort. Sepsis and PARDS were diagnosed according to the Phoenix and PALICC-2 criteria, respectively. Clinical data obtained within 2 hours of admission were analyzed. Correlation analysis, variance inflation factor assessment, and least absolute shrinkage and selection operator regression were followed by multivariable logistic regression to construct the final model. Performance was evaluated using the area under the receiver operating characteristic curve (AUC), Brier score, calibration assessment, and 1000 bootstrap resamples. An independent external cohort of 100 children, was used only to assess discrimination. RESULTS: Among 547 children, 146 (26.7%) developed PARDS within 48 hours. Lower hemoglobin (OR = 0.987, 95% CI: 0.976-0.997), higher LAR (OR = 1.548, 95% CI: 1.042-2.299), septic shock (OR = 4.158, 95% CI: 2.001-8.639), and a higher Phoenix Sepsis Score (OR = 3.438, 95% CI: 2.603-4.540) were independently associated with early PARDS. The model achieved an AUC of 0.911 (95% CI: 0.882-0.939) in the derivation cohort. The apparent and optimism-corrected C-indices were 0.911 and 0.908, respectively, and the corresponding Brier scores were 0.0968 and 0.0995. In the external cohort, the AUC was 0.873 (95% CI: 0.775-0.970). CONCLUSION: Elevated LAR, higher Phoenix Sepsis Score, septic shock, and lower HB levels were important risk-related factors for the development of PARDS in pediatric sepsis. Enhanced dynamic monitoring and early interventions should be implemented for high-risk patients to improve clinical outcomes.