Preliminary establishment and evaluation of prognostic warning scoring system for severe trauma

Abstract

Objective To explore the related risk factors of death in patients with severe trauma, and to establish a preliminary early warning scoring system for the prognosis of severe trauma (hereinafter referred to as the prediction model), so as to provide convenience for clinicians to judge the prognosis of patients early, timely and accurately. Methods The clinical data of 307 patients with severe trauma treated in the Emergency Department of the First Affiliated Hospital of Army Medical University from January 2019 to June 2021 were collected and retrospectively analyzed. They were divided into survival group (n=262) and death group (n=45) according to the outcomes. The age, sex, comorbidity, injury mechanism, number of injury sites, vital signs at admission, results of blood routine test, blood biochemistry test and blood coagulation test, blood gas analysis, endotracheal intubation at admission, emergency operation, use of vasoactive drugs and Glasgow coma score (GCS), injury severity score (ISS), and new injury severity score (NISS) were compared between the 2 groups with univariate analysis. The statistically significant indexes in univariate analysis were included in multivariate logistic regression analysis, and the risk factors affecting the prognosis of the patients were screened, and the prediction model was established according to the assignment of regression coefficient β, and receiver operating characteristic (ROC) curve was drawn to evaluate the predictive value of the prediction model for the prognosis of patients with severe trauma. Results There were differences in injury mechanism, number of injury sites, comorbidity, endotracheal intubation on admission and use of vasoactive drugs between the 2 groups. The age, serum sodium, serum creatinine (Scr), lactic acid, blood glucose, fibrin degradation products and D-dimer were higher, while the body temperature, blood potassium, activated partial thrombin time, oxygenation index and GCS were lower in the death group than the survival group. The results of multivariate logistic regression analysis showed that age (≥62.5 years), Scr (≥77.37 mmol/L), GCS (≤12.5), oxygenation index (≤253.5 mmHg) and body temperature (≤36.25 ℃) were independent risk factors for death in the patients with severe trauma (P < 0.05). The above 5 variables were scored to form the prediction model, which divided severe trauma patients into low-risk, medium-risk and high-risk, and the incidence of death was 2.2%, 28.2% and 74.4%, respectively. When the cut-off value of the prediction model was 15.5, the sensitivity was 86.7%, the specificity was 89.4%, and the Yoden index was 0.764. Conclusion The prediction model is composed of age, Scr, GCS, oxygenation index and body temperature. Its predictive effect is better than those of the 5 indicators alone, and has more advantages than ISS, NISS, acute physiology and chronic health evaluation Ⅱ, sequential organ failure assessment and modified early warning score. It is more helpful to evaluate the prognosis of patients with severe trauma quickly and accurately.

Keywords

• severe trauma
• prognosis
• prediction system
• establishment and evaluation