TRISS, NTRISS AND ASCOT VALIDATION IN SEVERE TRAUMA POPULATION ADMITTED IN MOLDOVIAN TRAUMA CENTER

Abstract

Introduction: Recent statistical data from different countries places the trauma among the leading causes of death with increasing contribution to the overall mortality rate. The possibility to predict and to anticipate the eventual complications could significantly increase trauma survival rate. This is possible by analyzing different clinical parameters of trauma patients and identifying those with high predictive power. The results were used to concept different traumatic scores. Nowadays, there are a lot of scores elaborated considering anatomical, physiological or mixed criteria. They were developed considering the particularities of distinct medical systems from different countries. They could differ in many aspects from each other, inclusively, from the Moldavian one. Thus, it is necessary to find the optimal score for daily use in distinct conditions for target population of current study. Purpose and task: Validation and comparative evaluation of the most common mixed traumatic scores in conditions of a trauma center from Republic of Moldova. Methods: In the current retrospective analytical research, was analyzed the data of 2651 severe trauma patient’s consecutive admitted in Moldavian trauma center in period between January 2013 – November 2018. The source for information was the electronic database of IMU with no personal information. The inclusion and exclusion criteria were respected. They were calculated ASCOT, TRISS and NTRISS scores to assess patient’s survival rate. The prediction results were compared and statistically analyzed by logistic regression. Results: The comparison of the mixed scores included in the research showed that the NTRISS score showed a maximum coefficient of determination compared to TRISS and ASCOT, all models having calibration indicators that need improvement, the criteria being the significance of the test. Conclusion: In this article, three common mixed predictive models were validated. Of these, NTRISS has optimal characteristics in terms of determination/ discrimination and could be recommended for daily use in conditions of a trauma center from Republic of Moldova.