Abstract: Objective To analyze the risk factors of postoperative hypothermia in patients undergoing craniocerebral surgery and construct a prediction model to provide reference for early clinical screening of high-risk groups and intervention. Methods By retrospective analysis method,407 patients from a tertiary Grade A hospital in Beijing from January 2021 to December 2022 were selected as the research objects,which were divided into modeling group（285 cases） and model testing group（122 cases） according to the ratio of 7：3. Univariate analysis and binary Logistic regression analysis were conducted in the modeling group and used to build the prediction model,and the prediction efficiency of the prediction model was evaluated. The data of model testing group was used to validate the model. Results The incidence of postoperative hypothermia in patients undergoing craniocerebral surgery was 40.7%. Logistic regression analysis showed that preoperative use of midazolam（OR=2.464）,operation time less than 3 hours（OR=3.287）,intraoperative use of colloidal fluid（OR=3.399） were independent risk factors for postoperative hypothermia in patients undergoing craniocerebral surgery. The prediction model was：Logit（P）=-2.124+0.902* preoperative use of midazolam +1.190* operative time less than 3 hours + 1.224* intraoperative use of colloidal fluid. The Hosmer-Lemeshow（H-L） test for the modeling group was（χ; ²=2.634,P=0.955）, the area under the ROC curve was 0.720. The H-L test for the modeling testing group was（χ; ²=13.911,P=0.084）,and the area under the ROC curve was 0.705. Conclusion The prediction model has good results and can provide reference for medical staff to screen high-risk groups with postoperative hypothermia in craniocerebral surgery patients.

Key words: Hypothermia, Craniocerebral Surgery, Prediction Model, Anesthesia Care