Abstract: Objective To analyze the risk factors of delirium in the elderly patients undergoing cardiac surgery,and to establish and validate a nomogram model for predicting the risk of delirium. Methods The elderly patients who underwent cardiac surgery and were admitted to intensive care unit (ICU) of the Second Affiliated Hospital of Hainan Medical College from January 2018 to August 2020 were enrolled. The combination of the Richmond Agitation Sedation Scale (RASS) and the ICU-Confusion Assessment Method (CAM-ICU) were used to evaluate delirium.The patients were divided into delirium group and non-delirium group according to whether delirium occurred within 3 days after ICU admission,and the clinical characteristics of the two groups were observed and compared. The risk factors of delirium were identified by multivariate Logistic regression analysis. The predicting nomogram model for delirium in the elderly patients undergoing cardiac surgery was established by R 3.6.2 software. The modelling was validated by the Bootstrap method.The deviation and prediction efficiency of nomogram model for delirium were evaluated by calibration curve and receiver operating characteristic(ROC) curve, respectively. Results A total of 274 cases were enrolled in this study. The incidence rate of delirium was 9.1%(25 cases).Multivariate Logistic regression analysis showed that advanced age(OR=1.038,95%CI:1.004-1.247), high Charlson Comorbidity Index score(OR=1.760,95%CI: 1.154-2.685), cardiopulmonary bypass(OR=5.489,95%CI: 1.501-20.068) and intraoperative hypotension(OR=4.630,95%CI: 1.260-17.015) were independent risk factors for delirium in the elderly patients undergoing cardiac surgery.The internal validation C-index of nomogram prediction model was 0.876(95%CI:0.799-0.952),with good discrimination and stability. The calibration curve revealed that the mean absolute error between predicted delirium risk and the actual delirium risk was 0.017,and ROC curve showed that the area under the curve(AUC) for the nomogram model predicting delirium was 0.871(95%CI:0.819-0.923). Conclusions The nomogram model established in this study for predicting the risk of delirium has good discrimination,accuracy and clinical practicability.And it has certain guiding significance for predicting high-risk population of delirium and making targeted prevention strategies.

Key words: cardiac surgery, intensive care unit, delirium, risk factor, nomogram, model