Background and purpose: Trauma scoring systems help physicians and nurses in recognizing the severity of trauma and its management. The purpose of this study was to investigate and compare the predictive ability of mortality rates of two trauma scoring systems (MGAP and GAP) in multiple trauma patients.
Materials and methods: In this cross-sectional study, we used the data available in electronic records for trauma patients admitted to Rasht Poorsina Hospital, Iran. Demographic data, and GAP and MGAP (Mechanism, Glasgow Coma Scale, Age, and Pressure) scores for 1541 patients were extracted. Then, the ability of both scores in short-term (24 hours) and long-term (4-week) mortality prediction rates was analyzed using SPSS V21.
Results: The surface areas under the curve ratio of ROC for predicting short-term mortality rates were 0.947 and 0.938, and for long-term mortality rates were 0.928 and 0.914 for GAP and MGAP, respectively (P<0.001).  Also, Kappa coefficient for agreement of both scoring systems was 0.754 and 0.462, respectively.  The best cut-off values for GAP and MGAP in predicting short-term mortality rates were 16 and 21 with sensitivity rates of 97.6% and 96.6%, specificity rates of 81.1% and 83%, and accuracy rates of 97.1% and 96.2%, respectively, while in the long- term mortality rates these values were 19 and 22 with sensitivity rates of 92.4% and 92.6%, specificity rates of 80%, and accuracy rates of 91.95% and 92.1%, respectively.
Conclusion: Both GAP and MGAP could appropriately predict mortality rate without any significant difference. Therefore, these scores could be used as triage tools, and in predicting the severity of injuries and mortality.