ABSTRACT
The Trauma and Injury Severity Score (TRISS) is a well-accepted model used to evaluate the quality of trauma care in the US. This research aims to study whether TRISS can be applied to evaluate trauma care and classify outcomes of road traffic injury patients in Thailand. A retrospective study was used to review the Thailand's Injury Surveillance System database from the 1st January to the 31st of December 1996. The study subjects were severe road traffic injury patients with blunt injuries. The TRISS model was applied to compute the survival probability for each patient. The chi-square goodness-of-fit was used to compare the survival probability distribution between the American Major Trauma Outcome of Study (MTOS) and the road traffic injuries in Thailand. The accuracy, sensitivity and specificity of the survival prediction by TRISS were evaluated. The distribution of survival probability between American trauma patients and Thai road traffic injury patients was significantly different (p-value < 0.00001). The TRISS model had high accuracy and sensitivity, but low specificity, in predicting the survival of Thai road traffic injuries. The MTOS and Thai road traffic injuries had different distributions for various factors such as the Revised Trauma Score (RTS), Injury Severity Score (ISS), and ages which effect injury survival. Due to these factors the distribution of survival probability between MTOS and Thai road traffic injuries was also significantly different. By applying TRISS, the survival prediction of Thai road traffic injuries resulted in a high number of false positives.