Journal of Trauma and Injury

The Korean Society of Traumatology (대한외상학회)
권호 표지
DOI QR코드

DOI QR Code

Predicting 30-day mortality in severely injured elderly patients with trauma in Korea using machine learning algorithms: a retrospective study

  • Jonghee Han (Department of Cardiovascular and Thoracic Surgery, Trauma Center, Chungbuk National University Hospital) ;
  • Su Young Yoon (Department of Cardiovascular and Thoracic Surgery, Trauma Center, Chungbuk National University Hospital) ;
  • Junepill Seok (Department of Trauma Surgery, Trauma Center, Chungbuk National University Hospital) ;
  • Jin Young Lee (Department of Trauma Surgery, Trauma Center, Chungbuk National University Hospital) ;
  • Jin Suk Lee (Department of Trauma Surgery, Trauma Center, Chungbuk National University Hospital) ;
  • Jin Bong Ye (Department of Trauma Surgery, Trauma Center, Chungbuk National University Hospital) ;
  • Younghoon Sul (Department of Trauma Surgery, Trauma Center, Chungbuk National University Hospital) ;
  • Se Heon Kim (Department of Trauma Surgery, Trauma Center, Chungbuk National University Hospital) ;
  • Hong Rye Kim (Department of Neurosurgery, Trauma Center, Chungbuk National University Hospital)
  • Received : 2024.04.22
  • Accepted : 2024.05.29
  • Published : 2024.09.30
Download PDF
⟨ Previous Next ⟩

Abstract

Purpose: The number of elderly patients with trauma is increasing; therefore, precise models are necessary to estimate the mortality risk of elderly patients with trauma for informed clinical decision-making. This study aimed to develop machine learning based predictive models that predict 30-day mortality in severely injured elderly patients with trauma and to compare the predictive performance of various machine learning models. Methods: This study targeted patients aged ≥65 years with an Injury Severity Score of ≥15 who visited the regional trauma center at Chungbuk National University Hospital between 2016 and 2022. Four machine learning models-logistic regression, decision tree, random forest, and eXtreme Gradient Boosting (XGBoost)-were developed to predict 30-day mortality. The models' performance was compared using metrics such as area under the receiver operating characteristic curve (AUC), accuracy, precision, recall, specificity, F1 score, as well as Shapley Additive Explanations (SHAP) values and learning curves. Results: The performance evaluation of the machine learning models for predicting mortality in severely injured elderly patients with trauma showed AUC values for logistic regression, decision tree, random forest, and XGBoost of 0.938, 0.863, 0.919, and 0.934, respectively. Among the four models, XGBoost demonstrated superior accuracy, precision, recall, specificity, and F1 score of 0.91, 0.72, 0.86, 0.92, and 0.78, respectively. Analysis of important features of XGBoost using SHAP revealed associations such as a high Glasgow Coma Scale negatively impacting mortality probability, while higher counts of transfused red blood cells were positively correlated with mortality probability. The learning curves indicated increased generalization and robustness as training examples increased. Conclusions: We showed that machine learning models, especially XGBoost, can be used to predict 30-day mortality in severely injured elderly patients with trauma. Prognostic tools utilizing these models are helpful for physicians to evaluate the risk of mortality in elderly patients with severe trauma.

Keywords

References

  1. DiMaggio C, Ayoung-Chee P, Shinseki M, et al. Traumatic injury in the United States: in-patient epidemiology 2000-2011. Injury 2016;47:1393-403. https://doi.org/10.1016/j.injury.2016.04.002
  2. The American Association for the Surgery of Trauma. Trauma facts [Internet]. The American Association for the Surgery of Trauma [cited 2024 Jan 2]. Available from: https://www.aast.org/resources/trauma-facts
  3. Zhao FZ, Wolf SE, Nakonezny PA, et al. Estimating geriatric mortality after injury using age, injury severity, and performance of a transfusion: the Geriatric Trauma Outcome Score. J Palliat Med 2015;18:677-81. https://doi.org/10.1089/jpm.2015.0027
  4. Korean Statistical Information Service (KOSIS). Emergency medical status statistics [Internet]. KOSIS; 2021 [cited 2023 Dec 22]. Available from: https://kosis.kr/statHtml/statHtml.do?orgId=411&tblId=DT_41104_244&conn_path=I2
  5. Sawada Y, Isshiki Y, Ichikawa Y, et al. The significance of the treatment for elderly severe trauma patients who required intensive care. Cureus 2023;15:e39110.
  6. Cook AC, Joseph B, Inaba K, et al. Multicenter external validation of the Geriatric Trauma Outcome Score: a study by the Prognostic Assessment of Life and Limitations After Trauma in the Elderly (PALLIATE) consortium. J Trauma Acute Care Surg 2016;80:204-9. https://doi.org/10.1097/TA.0000000000000926
  7. Boyd CR, Tolson MA, Copes WS. Evaluating trauma care: the TRISS method: Trauma Score and the Injury Severity Score. J Trauma 1987;27:370-8. https://doi.org/10.1097/00005373-198704000-00005
  8. Cruz JA, Wishart DS. Applications of machine learning in cancer prediction and prognosis. Cancer Inform 2007;2:59-77.
  9. Hou N, Li M, He L, et al. Predicting 30-days mortality for MIMIC-III patients with sepsis-3: a machine learning approach using XGboost. J Transl Med 2020;18:462.
  10. Xu Y, Han D, Huang T, et al. Predicting ICU mortality in rheumatic heart disease: comparison of XGBoost and logistic regression. Front Cardiovasc Med 2022;9:847206.
  11. Gao W, Wang J, Zhou L, et al. Prediction of acute kidney injury in ICU with gradient boosting decision tree algorithms. Comput Biol Med 2022;140:105097.
  12. Liu NT, Salinas J. Machine learning for predicting outcomes in trauma. Shock 2017;48:504-10. https://doi.org/10.1097/SHK.0000000000000898
  13. Hanko M, Grendar M, Snopko P, et al. Random forest-based prediction of outcome and mortality in patients with traumatic brain injury undergoing primary decompressive craniectomy. World Neurosurg 2021;148:e450-8. https://doi.org/10.1016/j.wneu.2021.01.002
  14. Pollard TJ, Johnson AE, Raffa JD, Celi LA, Mark RG, Badawi O. The eICU Collaborative Research Database, a freely available multi-center database for critical care research. Sci Data 2018;5:180178.
  15. Guo C, Pan J, Tian S, Gao Y. Using machine learning algorithms to predict 28-day mortality in critically ill elderly patients with colorectal cancer. J Int Med Res 2023;51:300060 5231198725.
  16. Choi H, Lee JY, Sul Y, et al. Comparing machine learning and logistic regression for acute kidney injury prediction in trauma patients: a retrospective observational study at a single tertiary medical center. Medicine (Baltimore) 2023;102:e34847.
  17. Rau CS, Kuo PJ, Chien PC, Huang CY, Hsieh HY, Hsieh CH. Mortality prediction in patients with isolated moderate and severe traumatic brain injury using machine learning models. PLoS One 2018;13:e0207192.
  18. Yang S, Cao L, Zhou Y, Hu C. A retrospective cohort study: predicting 90-day mortality for ICU trauma patients with a machine learning algorithm using XGBoost using MIMIC-III database. J Multidiscip Healthc 2023;16:2625-40. https://doi.org/10.2147/JMDH.S416943
  19. Muthukrishnan R, Rohini R. LASSO: a feature selection technique in predictive modeling for machine learning. 2016 IEEE International Conference on Advances in Computer Applications (ICACA); 2016 Oct 24; Coimbatore, India. IEEE; 2016. p. 18-20.
  20. Li Z. Extracting spatial effects from machine learning model using local interpretation method: an example of SHAP and XGBoost. Comput Environ Urban Syst 2022;96:101845.
  21. Allyn J, Allou N, Augustin P, et al. A comparison of a machine learning model with EuroSCORE II in predicting mortality after elective cardiac surgery: a decision curve analysis. PLoS One 2017;12:e0169772.
  22. Perlich C. Learning curves in machine learning. In: Sammut C, Webb GI, editors. Encyclopedia of machine learning. Springer; 2011. p. 577-80.
  23. Corbacioglu SK, Aksel G. Receiver operating characteristic curve analysis in diagnostic accuracy studies: a guide to interpreting the area under the curve value. Turk J Emerg Med 2023;23:195-8. https://doi.org/10.4103/tjem.tjem_182_23
  24. Hicks SA, Strumke I, Thambawita V, et al. On evaluation metrics for medical applications of artificial intelligence. Sci Rep 2022;12:5979.
  25. Muller AC, Guido S. Introduction to machine learning with Python. O'Reilly Media; 2016.
  26. Miao J, Zhu W. Precision-recall curve (PRC) classification trees. Evol Intell 2022;15:1545-69. https://doi.org/10.1007/s12065-021-00565-2
  27. Li DC, Liu CW, Hu SC. A learning method for the class imbalance problem with medical data sets. Comput Biol Med 2010;40:509-18. https://doi.org/10.1016/j.compbiomed.2010.03.005
  28. Ma Y, Gan M. Gradient boosting based prediction method for patient death in hospital treatment. In: Chen H, Zeng D, Yan X, Xing C, editors. Smart health. International Conference on Smart Health (ICSH) 2019; 2019 Jul 1-2; Shenzhen, China. Springer; 2019. p. 283-93.
  29. Wei C, Zhang L, Feng Y, Ma A, Kang Y. Machine learning model for predicting acute kidney injury progression in critically ill patients. BMC Med Inform Decis Mak 2022;22:17.
  30. Nohara Y, Matsumoto K, Soejima H, Nakashima N. Explanation of machine learning models using shapley additive explanation and application for real data in hospital. Comput Methods Programs Biomed 2022;214:106584.