Abstract: Objective: To develop and validate a risk prediction model for pressure injury in patients with traumatic brain injury, so as to provide a reference for clinical medical staff to accurately identify high-risk populations at an early stage. Methods: Patients with traumatic brain injury hospitalized in a tertiary grade A hospital in Lhasa between January 2023 and December 2024 were recruited using convenience sampling. A total of 801 eligible cases were randomly assigned to a training set and an internal validation set at a 7 : 3 ratio (561 and 240 cases, respectively), and an additional 135 cases (January to July 2025) were included for external validation. Pressure injury risk was evaluated using the Braden Scale. Univariate and multivariate Logistic regression analysis were conducted to develop the prediction model, and model performance was assessed by means of Receiver Operating Characteristic (ROC) curves and calibration curves. Results: The incidence of pressure injury in training set was 8.4%. Age, blood glucose, state of consciousness, BMI, no preventive measures were taken, ICU length of stay, and ventilator days were identified as influencing factors for pressure injuries in patients with traumatic brain injury (all P<0.05). In the training set, the Area Under the Curve (AUC) was 0.840, with a sensitivity of 82.3% and a specificity of 74.5%. In the internal validation set, the AUC was 0.929, with a sensitivity of 82.0% and a specificity of 95.5%. In the external validation set, the AUC was 0.990, with a sensitivity of 92.9% and a specificity of 100.0%. The Hosmer-Lemeshow test result of the training set was χ2=4.463, P=0.813. Conclusion: The risk prediction model developed in this study is highly accurate and practical. It helps clinical medical staff identify high-risk patients, implement targeted preventive and therapeutic measures, and reduce the incidence of pressure injuries.

Key words: traumatic brain injury; pressure injury; influencing factor; predictive model