Prediction Model for Survival of Trauma Patients

semanticscholar(2015)

引用 0|浏览1
暂无评分
摘要
One of the most common calculators used for predicting survival of trauma patients is TRISS (Trauma Injury Severity Score) which is based on multiple regression analysis from the Major Trauma Outcome Study (MTOS) database. Based on the data set of 1,009 trauma patients of a tertiary care hospital in Pakistan, data mining was carried out using Logistic Regression Modelling using an all subsets approach. Models were formulated and ranked based on accuracy, Area under Receiver Operating Curve (AUC) and an overall performance measure that include AUC, sensitivity and specificity. Various models were found to perform better than TRISS. Furthermore, reduced models were also formulated that include factors that are simple to determine and models were found to outperform TRISS.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要