Document Type : Original Article
Authors
1 Department of Biostatistics and Epidemiology, Faculty of Health, Tabriz University of Medical Sciences, Tabriz, Iran
2 Road Traffic Injury Research Center, Tabriz University of Medical Sciences, Tabriz, Iran
3 Department of emergency, school of medicine, Ardabil university of medical sciences, Ardabil, Iran
Abstract
Objective: Elderly people now live longer and are more active, but their weak bodies make them more susceptible to injuries. A suitable tool to estimate mortality risk is required for older, injured patients. To fill a gap in existing general models, this study intends to develop a modified model based on injury severity to predict trauma-related mortality in older people.
Methods: The study analyzed data from 683 older trauma patients aged 55 and above from Fatemi University Hospital in Ardabil between 2019 and 2022. It used regression modeling to examine the association between predictor variables and death. A Geriatric Injury Prognostic (GIP) model was constructed and evaluated for its efficacy in distinguishing mortality statuses using the AUC, odds ratio, and Hosmer-Lemeshow test, with significance tested at a 0.05 level, and its internal and external validity was assessed using the AUC and ROC curves.
Results: The study enrolled 643 patients with a trauma age range of between 55 and 95 years, with 62.36% being men. In the emergency department, the mean pulse rate was 81.7 ± 8.2, 5% of patients had GCS less than 13, the mean ISS score for 636 patients was 8 ± 8.5, and in-hospital mortality was 6.22%. Using pulse rate per minute, the Glasgow Coma Scale (GCS), and the Injury Severity Score (ISS), a comprehensive multivariate model was developed, effectively predicting 83.16% of trauma-related mortality in the older population. The GIP model's area under the rock curve (AUC) value was 0.995.
Conclusion: The GIP model accurately predicts death probability in the older population, aiding in decision-making for appropriate treatment approaches in the geriatric-specific population.
Keywords
Main Subjects