Rapid Identification of Critical Factors Influencing Pediatric Trauma Mortality in Disaster Settings

Introduction: This study aimed to investigate the early assessment of pediatric trauma patients(PTPs) following earthquakes, identifying key in-hospital mortality factors. Effective early assessments and interventions were crucial for pediatric patients due to their unique physiological and psychological responses. Aligning with the Sendai Framework for Disaster Risk Reduction 2015-2030, this research focused on reducing disaster-related mortality among vulnerable populations like children. Prior research had focused on advanced tools, often found to be infeasible post-earthquake. Furthermore, multicollinearity among factors complicated statistical analysis, thus necessitating the identification of simple, rapid assessment indicators that predicted pediatric mortality. Methods: This study undertook a comprehensive retrospective analysis of 200 PTPs recorded in the Western China Earthquake Patient Database(WCEPD). The study employed Least Absolute Shrinkage and Selection Operator(LASSO) regression to identify independent variables influencing in-hospital mortality. The study collected 20 baseline indicators that could be swiftly obtained upon patient admission. These indicators included demographic information, vital signs, consciousness status, and injury specifics, such as the nature and location of injuries. Results: The LASSO regression analysis identified 5 independent factors significantly associated with in-hospital mortality among PTPs: Glasgow Coma Scale/Pediatric Glasgow Coma Scale(GCS/PGCS), Systolic Blood Pressure(SBP), Respiratory Rate(RR), Pulse Rate(PR), and crush injury. Based on these factors, a predictive nomogram was constructed to identify PTPs with elevated in-hospital mortality risks. Conclusion: Through the application of LASSO regression, this study successfully identified and analyzed 5 simple, rapidly assessable factors predictive of pediatric trauma mortality in disaster settings. The development of a predictive nomogram incorporating these factors provided a powerful tool for early identification and timely intervention.