Postoperative Delirium (POD) is a common complication in patients after cardiac surgery, which is associated with short-term prognosis and long-term quality of life. Due to the lack of specific treatment, prevention is the most effective strategy to reduce delirium. Risk assessment greatly contributes to guide prevention by stratifying the risk of the POD. However, the high-quality risk assessment tools are still sparse, and the combination of risk assessment and stratified prevention has not been applied to the management of delirium after cardiac surgery. In the previous work, we found the predictive value of preoperative cardiac function on POD, and found a critical value. On the basis of previous research, this project will learn from previous studies on risk factors of POD, and deeply explore potential predictors. Furthermore, we will develop and validate POD risk prediction model, and then develop it into a visual evaluation system. In addition, based on the theory of risk management and risk assessment tools, using the concepts and methods of evidence-based medicine, the risk assessment system and stratified management program of POD will be formed and evaluated by expert discussion meetings. It is expected that the risk assessment system and stratified management program constructed by this project could simply and quickly screen high-risk patients and carry out timely intervention, so as to reduce the incidence of POD, improve patients' prognosis and the quality of life after cardiac surgery.

Inclusion Criteria:

• Aged above 18
• Patients after cardiac surgery
• Admitted to ICU after surgery

Exclusion Criteria:

• Preoperative delirium or dementia patients
• Unable to fully participate in delirium testing, including blind, deaf, illiterate or inability to understand Chinese
• Undergoing surgery procedures do not require admission to SICU
• Transfer to ICU from wards after surgery

Patients aged above 18 who are admitted to surgical ICUs after cardiac surgery will be screened for eligibility.