Abstract:Objective To preliminarily construct a Hypothermia Risk Score for traumatic patients admitted to emergency department, and verify its clinical predictive efficacy in clinical setting. Methods The clinical data of 489 trauma patients admitted to the emergency department were collected, and served as derivative group. According to the occurrence of hypothermia, the trauma patients were divided into a hypothermia group and a normothermia group. The independent risk factors influencing hypothermia in trauma patients were analyzed using multivariate logistic regression analysis, based on which, the Hypothermia Risk Score was constructed.Then the cut-off value of the instrument was determined using ROC curve. Additionally, another 228 trauma patients admitted to emergency department were selected as the validation group, who was assessed with the Hypothermia Risk Score, and the assessment results were compared with de facto occurrence of hypothermia. Results The Hypothermia Risk Score derived from the derivative group consisted of 4 items: environmental temperature ≤8℃(scoring 1 point), RTS <4(scoring 2 points), damp clo-thing (scoring 2 points), and shock at admission (scoring 2 points). In the validation group, the assessment results of Hypothermia Risk Score had significant difference between the hypothermia subgroup and the normothermia group (P<0.01).At 3 points, the sensitivity, specificity, positive predictive value, negative predictive value of the instrument stood at 0.709, 0.920, 82.24%, and 85.63% respectively, with the overall correct rate of 86.08%,and the area under the ROC curve standing at 0.829[95%CI(0.769,0.888)]. Conclusion The Hypothermia Risk Score has a high differentiation and predictive value.It can be used to identify trauma patients at high risk of hypothermia after trauma.