Abstract: Objective: To systematically review the risk prediction model of Inadvertent Perioperative Hypothermia (IPH). Methods: A literature search was performed in the PubMed, Embase, Cochrane Library, CHINAL, Web of Science, CNKI, Wanfang, VIP and Chinese biomedical database to identify IPH risk prediction model relevant studies from inception of databases to November 1, 2021. Two researchers independently screened the literatures, extracted the data and evaluated the quality of the included studies based on PROBAST tool. Results: Totally 12 literatures were included, including 13 models. BMI, baseline temperature, age, operating room temperature, fluid volume, anesthesia time and operation time were the main predictors of IPH model. The area under the receiver operating characteristic curve (AUC) of 10 models were >0.7 (0.789-0.936) in the modeling population, and the AUC of 3 models in the testing population were >0.7 (0.771-0.914). One model was internally validated and seven models were external validated. All studies had certain bias risks, but the overall applicability was relatively good. Conclusion: The IPH risk prediction model has good predictive performance, but high risk of bias. Existing models should be prudently selected for further validation, or we could carry out large sample and multiple disease prospective clinical studies to construct native IPH optimal risk prediction model, in order to identify and prevent IPH as early as possible.

Key words: Inadvertent Perioperative Hypothermia; predication; model; systematic review