Abstract: Objective To evaluate the effects of a risk prediction model for postoperative delirium in ICU patients after surgery. Methods A total of 240 ICU patients in a tertiary hospital were selected, and indicators of ICU postoperative delirium (ICU-P0D) group n=49) and non ICU-P0D group n=191) were compared. Logistic regression was used to establish a risk prediction model. The area under the R0C curve was used to test the model to predict the effects. Twenty-five patients were selected to evaluate the model’s effects. Results The study finally included POSSUM score (OR=4.259), acid-base imbalance (OR=28.731), diabetes (OR=2.699), hypertension (OR=2.055),and transient coma before admission(OR=10.719) to construct the risk prediction model. The model formula was:Z=0.121 × POSSUM + 2.519 × transient coma before admission +1.157 × hypertension + 1.004 × diabetes + 0.26 × acid-base imbalance -6.957,the area under the ROC curve of this model was 0.832,with the sensitivity of 0.639,the specificity of 0.886, the youden of 0.526. The model verification results showed the sensitivity of 60%，the specificity of 100%，and the accuracy of 92%. Conclusion The risk prediction model has satisfactory prediction effects. The risk prediction can be completed at the beginning of ICU stay,which can provide reference for preventative treatment and nursing measures for high-risk patients.

Key words: Intensive Care Units, Delirium, Risk Prediction Model

摘要： 目的 探讨ICU术后谵妄风险预测模型预测患者发生ICU术后谵妄的效果。方法 选取某三级甲等医院术后进入ICU治疗的患者240例，将ICU术后谵妄组(n=49)和非谵妄组（n=191)的各项指标进行对比，利用Logistic回归建立风险预测模型，应用ROC曲线下面积检验模型预测效果，并选取25例患者进行模型预测效果验证。结果 本研究最终纳入病死率及并发症发生率的生理学和手术严重性评分系统(OR=4.259)、酸碱失衡(OR=28.731)、糖尿病(OR=2.699)、高血压(OR=2.055)、入院前一过性意识丧失（OR=10.719)5个危险因素构建出风险预测模型，模型公式:Z=0.121×用于计数病死率及并发症发生率的生理学和手术严重性评分系统的评分+2.519×入院前一过性意识丧失的赋值+1.157×高血压的赋值+1.004×糖尿病的赋值+0.260×酸碱失衡评分-6.957。本模型的R0C曲线下面积为0.832,灵敏度为0.639,特异度为0.886，Youden指数为0.526。模型验证结果：灵敏度为60%，特异度为100%，准确率为92%。结论 本模型预测效果良好，适用于临床实际，患者进入ICU之初即可预测ICU术后谵妄的发生风险，为临床医护人员及时对高危患者采取预防性治疗和护理提供参考。

关键词: 重症监护病房, 谵妄, 风险预测模型