Construction And Verification Of Prediction Model Of Delirium Risk In Elderly Patients After Open Heart Surgery

Objective To explore the risk factors of delirium in patients after cardiac surgery,construct a risk prediction model of delirium for elderly patients after open heart surgery,and to verify the prediction effect of the model.Methods A single-center prospective cohort study was conducted.Through literature review and expert meeting,19 prediction factors were established.Using the convenient sampling method,prospective patients who were admitted to the Cardiac surgical ICU of a third-class A general hospital from February 2022 to August 2022 and who met the standard were selected as the research objects.The anxiety and depression were investigated by the research team members before operation.The presence of delirium in the Cardiac surgical ICU was evaluated every 12h(8:00/20:00)by the well-trained and qualified bedside nurses using the Confusion Assessment Method for the ICU(CAM-ICU).70%cohort was used for model development and the remaining 30%for internal validation.The logistic regression method was used to develop the prediction model.Single factor analysis is used for the preliminary screening of candidate factors.In univariate analysis,the candidate factors with statistically significant differences(P<0.05)entered multivariate Logistic regression to finally determine the independent predictors of delirium in elderly patients after open heart surgery.Hosmer-Lemeshow test was used to judge the goodness of fit of the model,and the area under receiver operating characteristic curve was used to test the model discrimination,and the prediction accuracy was calculated.Results A total of 266 SCICU patients were selected this study,and finally 249 patients who met the standard were included.Among them,there were 175 cases in the modeling group and 74 cases in the verification group,and the incidence of delirium was 23.69%.Univariate analysis showed that the influencing factors of delirium in elderly patients after cardiac surgery were age,hypertension history,cardiopulmonary bypass(CPB)time,aortic occlusion time,blood loss,mechanical ventilation time,ICU stay time,hypoxemia,APACHE-Ⅱ score,HAD score,lactic acid,and use of midazolam.There were significant differences between the two groups(P<0.05).Multivariate Logistic regression analysis showed that age,aortic occlusion time,hypoxemia,HAD score,and use of midazolam were independent risk factors.The area under the ROC curve of this model was 0.879(95%CI:0.815～0.943,P<0.001),with the sensitivity of 0.804,the specificity of 0.837 and the youden of 0.641.The area under the ROC curve of model verification was 0.903(95%CI:0.824～0.981,P<0.001),the sensitivity specificity and accuracy were 69.23%,90.16%and 86.49%,respectively.Conclusion This risk prediction model provides an effective prediction effect,and it is simple and convenient to use in clinical application,which could be a reference for medical staff to take preventive treatment and nursing care.