Introduction: Earthquakes pose a significant threat to public health, particularly among the elderly population, who are more susceptible to injury and have a higher prevalence of comorbidities. Traditional disaster triage protocols may not adequately account for the complex interplay between acute trauma and pre-existing health conditions. This study aims to refine the triage process for elderly earthquake victims by integrating comorbidities into the assessment. Methods: We conducted a retrospective analysis of a large dataset comprising 25,006 trauma patients aged 60 and older, extracted from public databases. The dataset included information on patient demographics, pre-existing comorbidities (such as hypertension, coronary heart disease, cancer, and diabetes), post-injury vital signs, and outcomes (survival or death). To identify the most significant predictors of mortality among these patients, we utilized the Least Absolute Shrinkage and Selection Operator (LASSO) regression technique. We applied LASSO regression with cross validation to determine the optimal penalty parameter (lambda) that minimizes the mean squared error and ensures the model’s generalizability. Variables with non-zero coefficients in the LASSO model were deemed significant predictors of mortality. Using these predictors, we developed a nomogram—a visual, user-friendly tool—to assist healthcare professionals in making informed triage decisions in the chaotic aftermath of an earthquake. Results: The LASSO regression analysis revealed renal failure, coronary heart disease, and the Revised Trauma Score (RTS) as the most influential predictors of mortality in the elderly trauma patients (p < 0.05). The nomogram incorporating these factors was validated and shown to have good discriminative ability and calibration. Conclusion: The integration of LASSO regression and nomogram development offers a statistically rigorous and practical approach to enhancing the accuracy and efficiency of triage decisions for elderly earthquake victims, potentially improving survival rates and resource allocation in disaster response scenarios.
Hai Hu (Sun,) studied this question.