Methods: Retrospective cross-sectional design was applied. Emergency department visits from 2008 to 2018, Jane where traumatic brain injury was evaluated and diagnosed in a medical center of Northern Taiwan. Decision tree was used to express a sequential classification and describe by a set of attributes. Decision tree techniques have been widely used to build classification models as such models closely resemble human reasoning and are easy to understand. Decision trees are sequential models, which logically combine a sequence of simple tests; each test compares a numeric attribute against a threshold value or a nominal attribute against a set of possible values.
Results: Several decision trees algorithms were used in this research. The algorithm that produced the highest accuracy was chosen as the most successful algorithm for modeling the mortality rate prediction. The highest classification rate of 94.7% was produced using the Random Forest decision tree algorithm. Therefore, the most significant variable was motor performance by assessing Glasgow coma scale.
Conclusion: It has been shown that data mining tools can largely be used by clinical professionals for predicting outcome of nursing care and mortality. High accuracy classification rate of mortality rate prediction among TBI patients in the decision tree algorithms, guiding the clinical health professionals in interesting treatment directions and shared decision making with family members of traumatic brain injury patients.