Abstract: Objective To investigate the incidence and risk factors of in-hospital hypothermia in the elderly patients with severe trauma at Emergency Department, and to develop and validate a prediction model. Methods A retrospective analysis was conducted on 524 elderly patients with severe trauma admitted to Emergency Department of the Second Affiliated Hospital of Naval Medical University from September 2020 to September 2024. Patients were randomly divided into a modeling cohort and a validation cohort at a ratio of 7∶3. According to the occurrence of hypothermia, each cohort was further stratified into a hypothermia group and a non-hypothermia group. After univariate analysis of candidate variables, multivariate logistic regression was performed to identify independent risk factors. A nomogram-based prediction model for in-hospital hypothermia in the elderly patients with severe trauma was constructed accordingly. The model’s discrimination, calibration, and clinical applicability were evaluated by the receiver operating characteristic(ROC) curve, calibration curve, and decision curve analysis (DCA), respectively. Results Of the 367 patients in the modeling cohort, 65 developed hypothermia, with an incidence rate of 17.71% (65/367). Multivariate logistic regression identified core body temperature (CBT) on arrival, injury severity score (ISS), hemoglobin (Hb), fibrinogen (FIB), and lactate (Lac) were independent risk factors for in-hospital hypothermia (P<0.05). The area under the ROC curve (AUC) of the prediction model was 0.874(95%CI:0.782-0.945)in the modeling cohort and 0.863(95%CI: 0.721-0.914) in the validation cohort, which indicated good discrimination. Calibration curves showed close agreement between the predicted probabilities and actual occurrence of hypothermia in both datasets, demonstrating good calibration. DCA indicated substantial net benefit, supporting favorable clinical utility. Conclusions CBT, ISS, Hb, FIB, and Lac are independent risk factors for in-hospital hypothermia in the elderly patients with severe trauma at Emergency Department. The prediction model constructed based on these factors shows high discrimination and good clinical application value.

Key words: emergency department, severe trauma, aged, in-hospital hypothermia, prediction model