Abstract: Objective To explore the risk factors of craniofacial medical device-related pressure injury in ICU patients and to establish a predictive model. Methods Two hundred and ten ICU patients from June 2021 to February 2022 were divided into group A (n=50, with craniofacial medical device-related pressure injury) and B (n=160, without craniofacial medical device-related pressure injury). The binary logistic regression model was used to analyze the risk factors, and a prediction model was constructed and verified. Results Among the 210 subjects, medical device-related pressure injury was observed in 50 patients (23.8%). Multivariate Logistic regression analysis showed that edema, prone position, acute physiology and chronic health evaluationⅡ score and time of using vasoactive drugs were independent factors affecting the craniofacial medical device-related pressure injury of ICU patients. A regression equation was established to predict the risk of craniofacial medical device-related pressure injury:Logit(P)=12.399+2.153×edema+2.217×prone position+0.541×APACHEⅡscore+0.342×vasoactive drug use time. AUC of nomogram model was 0.945（95%CI：0.926~0.964）, and the model differentiation was good. The calibration curve showed that the calibration degree of the model was better, and the Hosmer-Lemeshow test（χ²=3.063, P=0.930）showed that the model fitted well. Conclusion The craniofacial medical device-related pressure injury risk prediction model of ICU patients constructed in this study has a good risk identification ability and can provide reference for early screening of high-risk patients.

Key words: craniofacial, medical device-related pressure injury, predictive model, nomogram