Design a Model to Predict Hospital Mortality of Multiple Trauma Patients caused by Traffic Accidents

Abstract

The aim of this study was to design a model to predict hospital mortality of multiple trauma patients caused by traffic accidents. Materials and Methods: This study was cross-sectional study and performed in 2019 to 2020 in the pre-hospital emergency center of Tabriz and Imam Reza and Shohada Hospital of Tabriz, which are the referral centers for trauma patients. The sample size of this study was estimated to be 600 people. The sampling method in this study was easy and available sampling based on inclusion and exclusion criteria. To conduct the study after explaining the study to the participants and their companions of one degree and obtaining informed consent, information about age, sex, Respiratory Rate, O2 saturation, heart rate, initial blood pressure, primary GCS, level of consciousness based on AVPU, trauma mechanism, the type of vehicle involved in the accident, the location of the injured person in the vehicle, the location of the accident, the hospital outcome including mortality in the emergency department, mortality in the collected ward and the final diagnosis of the patients were recorded. The final outcome was recorded as death or alive at the time of discharge from the hospital. Patients' outcome was also recorded using the GOS (Glasgow Outcome Scale) system. Based on the obtained data, GAP, RTS, NTS scores were also collected and their results were compared with the designed model. Results: In this study, the mean (standard deviation) age of the subjects was 34.50 (±16.7) years. The highest outcome of patients based on GOS system was recovery, severe disability and moderate disability with 243 cases (40.5%), 162 cases (27.0%) and 158 cases (26.3%), respectively. The mean (SD) of GAP score in the subjects was 21.89 (±3.2) and the severity of GAP in most subjects in this study was low with 543 cases (90.5%). Also, the mean (SD) of NTS score in the subjects was 21.56 (±3.1) and the severity of NTS in most subjects in this study was low with 552 cases (92.0%). The mean (SD) RTS score in these subjects was 7.59 (±0.8). Among the significant variables included in the regression model, the primary GCS variables (P-value = 0.010), GAP (P-value = 0.003), NTS (P-value = 0.022) and AVPU (P-value <0.001) with adjustment on other variables significantly the odds os mortality predicted 0.326, 1.929, 1.965 and 0.168 times, respectively; On the other hand, there was no significant relationship between mortality and other significant variables (P-value> 0.05).