两种风险预测模型在预测ICU术后谵妄中的效果比较

目的 探讨早期谵妄风险预测模型和术后谵妄风险预测模型在预测ICU术后患者发生谵妄风险的效果.方法 选取2020年3—12月在某三甲医院外科术后入住ICU继续接受治疗的患者为研究对象,在患者入住ICU时分别收集两个风险预测模型所需的数据,应用重症监护谵妄筛查量表(intensive care delirium scree...

Comparison of the effectiveness of two risk prediction models in predicting postoperative delirium of ICU patients

Objective To compare the effect of the early prediction model for delirium and postoperative delirium risk prediction model in predicting the risk of postoperative delirium of ICU patients.Methods Patients admitted to ICU between March 2020 and December 2020 after surgery in a grade-3 class-A hospital were selected as the research subjects.Data required by the two risk prediction models were collected when the patients were admitted to ICU,and the intensive care delirium screening checklist(ICDSC)was used to evaluate postoperative delirium of the patients.The area under receiver operating characteristic curve(AUROC)was used to compare the prediction effects of the two models.Results A total of 610 patients were enrolled,of which 194 patients developed postoperative delirium in ICU,with an incidence of 31.80%.The AUROC of the early-stage delirium risk prediction model was 0.750(95%CI:0.687~0.813)with an accuracy of 80.98%,while that of the postoperative delirium risk model was 0.834(95%CI:0.781~0.888),with an accuracy was 87.21%.Significant differences were found in AUROC between the two models(Z=2.427,P=0.015).Conclusion The postoperative delirium risk prediction model is superior to the early prediction model for delirium.Therefore,it is suggested that ICU nurses should give priority in using the postoperative delirium risk prediction model.