Document Type : Original Article
Authors
1 Department of Emergency Medicine, Faculty of Medicine, Universitas Brawijaya, Saiful Anwar General Hospital, Malang – Indonesia
2 Department of Emergency Medicine, Faculty of Medicine, Universitas Brawijaya, Saiful Anwar General Hospital, Malang – Indonesia
3 Department of Public Health, Faculty of Medicine, Universitas Brawijaya, Malang - Indonesia
Abstract
Objective: Trauma-related deaths are among the top 10 causes of mortality, with an average of three deaths from traffic accidents every hour in Indonesia. In 2012, there were 117,949 traffic accidents resulting in 29,544 deaths (25.04%). In 2021, there were 103,645 accidents with 25,266 deaths (24.37%). Despite efforts to record trauma cases in Indonesia, existing scoring systems from developed countries face limitations. This study aims to propose a new, locally adapted scoring system to improve the management of multiple trauma cases, particularly at Saiful Anwar General Hospital (RSSA), Malang.
Methods: This observational analytic study with a retrospective cohort design was conducted at RSSA, Malang, Indonesia, from January 2021 to December 2022. A total of 506 multiple trauma patients from the RSSA Emergency Department were included, selected through purposive sampling. Data analysis involved the use of the t-test or the Mann-Whitney U test for numerical and ordinal data, the chi-square or Fisher's test for nominal data, followed by multivariate logistic regression to establish a scoring system.
Results: Logistic regression through backward elimination identified 15 significant predictors of in-hospital mortality: age (p = 0.000, OR = 0.967), pulse rate (p = 0.006, OR = 0.981), GCS (p = 0.000, OR = 1.381), intracerebral hemorrhage (p = 0.105, OR = 0.966), subdural hemorrhage (p = 0.001, OR = 0.875), infratentorial hemorrhage (p = 0.000, OR = 0.151), subfalcine herniation (p = 0.038, OR = 0.871), transtentorial herniation (p = 0.050, OR = 0.038), mandibular fracture (p = 0.004, OR = 0.235), etc. GCS was the strongest predictor (Wald = 50.54). Although intracerebral hemorrhage and lung tissue injury (p > 0.05) were retained due to clinical relevance, the model showed excellent discriminatory power, with an area under the curve (AUC-ROC) of 0.917 and a 95% confidence interval of 0.886–0.948.
Conclusion: The MTPS scoring system developed in this study can predict the prognosis of multiple trauma patients with strong discrimination (0.917) and is expected to improve the quality of trauma care in accordance with WHO guidelines at RSSA Malang.
Keywords
Main Subjects
)