Objective:To construct the prognostic prediction model and scoring tool by using severe trauma patients’ physiological indicators on admission, and to verify the clinical application effect and provide a reference for the early evaluation of severe trauma patients.Methods:This study was a retrospective study which adopted cluster sampling. Patients who met the inclusion and exclusion criteria in the emergency department of the First Affiliated Hospital of Soochow University from September 2019 to November 2020 were included. Patients were randomly assigned into the modeling group and the validation group in a ratio of 7:3 based on their outcome in the emergency department. Logistic regression analysis was performed to construct a prediction model, which was simplified as a scoring tool. The model was verified by using validation group and two months’ prospective validation. The efficiency of the simplified scoring tool was compared with that of the revised trauma score (RTS) and the injury severity score (ISS).Results:Totally 863 patients were included in this study, including 604 patients in the modeling group and 259 patients in the validation group. The model included systolic blood, SpO 2 and AVPU score. The AUC for predicting the death of severe trauma patients was 0.938. The AUC of the prediction model was 0.933, the best cut-off point was 5, the sensitivity was 86.7%, the specificity was 94.2%; the AUC of the validation was 0.885, the sensitivity was 83.3%, the specificity was 93.7%; and the AUC of prospective validation was 0.919, the sensitivity was 100%, and the specificity was 76.7%. The AUC of the RTS and ISS were 0.800 and 0.833, respectively. The AUC of RTS was lower than that of the simplified scoring tool constructed in this research. Conclusions:The prediction model and simplified scoring tool are better than RTS in predicting the outcome of emergency severe trauma patients, which are convenient for emergency medical staff to evaluate the severity of trauma patients.