Abstract: Objective To analyze the risk factors of pressure injury (PI) in ICU patients and establish a nomogram model for predicting the risk of PI. Methods A total of 341 patients admitted to ICU from January 2017 to December 2017 who met the inclusion criteria were selected as research subjects. According to the occurrence of pressure sores in patients, the clinical data of patients during hospitalization were retrospectively collected by a self-designed questionnaire. Logistic regression analysis was used to determine independent risk factors of PI. Nomogram model for predicting the occurrence of PI was established. H-L deviation test and ROC curve were used to evaluate the deviation and discrimination of the prediction model respectively. Results The incidence of PI in this study was 16.4%. Multivariate logistic regression analysis showed that stroke, diabetes, lactic acid, invasive mechanical ventilation, gender, hemoglobin and serum albumin were independent influence factors for the occurrence of PI（all P<0.05）. The established nomogram model had good accuracy (H-L test: χ²=4.561, P=0.683) and discrimination (AUC=0.886, 95%CI：0.749~0.970). Conclusion This study integrates 7 independent influence factors of PI, and the corresponding nomogram prediction model has good predictive value, which is helpful for medical personnel to adopt targeted prevention strategies, thus to effectively prevent the occurrence of PI in ICU patients.

Key words: severe patient, pressure injury, prediction model, the column chart