This study aimed to investigate spontaneous hypothermia among emergency trauma patients and develop a predictive model. A cohort of 162 emergency trauma patients was categorized into hypothermic (n = 61) and control (n = 101) groups, with trauma severity assessed using the modified Glasgow Coma Scale (GCS). Univariate analysis revealed significant differences between the groups in trauma severity, posture, garment wetness, warming measures, pre-hospital fluid resuscitation, and modified GCS scores (P < 0.05). The hypothermic group exhibited lower prothrombin time compared to the control group (P < 0.05). A logistic regression model was constructed, expressed as Y = 25.76 - 1.030X 1 + 0.725X 2 + 0.922X 3 - 0.750X 4 - 0.57X 6, and its fit was evaluated using the Hosmer-Lemeshow test. The receiver operating characteristic curve demonstrated an area under the curve of 0.871, with 81.2% sensitivity and 79.5% specificity. The Youden index identified the optimal predictive cut-off at its highest (0.58). Validation results included 86.21% sensitivity, 82.93% specificity, and 84.29% accuracy. Risk factors for spontaneous hypothermia after emergency trauma encompassed trauma severity, posture during consultation, clothing dampness upon admission, warming measures during transfer, pre-hospital fluid resuscitation, and modified GCS scores. The risk prediction model demonstrated high accuracy, enabling effective assessment of spontaneous hypothermia risk in emergency trauma patients and facilitating preventive measures.
Read full abstract