Construction and Application of a Model for Predicting the Risk of Delirium in Postoperative Patients With Type a Aortic Dissection

Background: Postoperative delirium (POD), an alteration in a patient's consciousness pattern, can affect the treatment and prognosis of a disease. Objective: To construct a prediction model for delirium in patients with type A aortic dissection after surgery and to validate its effectiveness. Methods: A retrospective cohort design was used to study 438 patients undergoing surgical treatment for type A aortic dissection from April 2019 to June 2020 in tertiary care hospitals. POD (n = 78) and non-delirium groups (n = 360) were compared and analyzed for each index in the perioperative period. A prediction model was established using multifactorial logistic regression, and 30 patients' perioperative data were collected for model validation. Results: Eight predictors were included in this study: smoking, diabetes, previous cardiovascular surgery, ejection fraction (EF), time to aortic block, acute kidney injury, low cardiac output syndrome, and pulmonary complications. The area under the receiver operating characteristic (ROC) curve of the constructed prediction model was 0.98 ± 0.005, and the Youden index was 0.91. The validation results showed 97% sensitivity, 100% specificity, and 93% accuracy. The expression of the model was Z = Smoking assignment(*) – 2.807 – 6.009(*)Diabetes assignment – 2.994(*)Previous cardiovascular surgery assignment – 0.129(*)Ejection fraction assignment + 0.071(*)Brain perfusion time assignment – 2.583(*)Acute kidney injury assignment – 2.916(*)Low cardiac output syndrome assignment – 3.461(*)Pulmonary related complications assignment + 20.576. Conclusion: The construction of an effective prediction model for the risk of delirium in patients after type A aortic stratification can help identify patients at high risk of POD early. It also provides a reference for healthcare professionals in the prevention and care of these patients.