综合ICU患者获得性衰弱风险预测模型的构建与应用

Establishment and application of a risk prediction model for ICU acquired weakness

目的通过分析综合ICU患者获得性衰弱的危险因素,构建风险预测模型,并验证模型的应用效果。方法选取2018年11月至2019年10月在江苏大学附属医院综合ICU治疗的247例患者,将其分为ICU获得性衰弱组(n=106)和非ICU获得性衰弱组(n=141),将2组的各项指标进行对比并应用二元Logistic回归构建预测模型,采用H-L判断模型的拟合度,采用ROC曲线下面积检验模型预测效果。于2019年11月至2020年5月纳入106例患者作为验证组对模型进行临床应用效果的验证。结果本研究建模组ICU获得性衰弱发生率为42.91%(106/247),验证组为44.34%(47/106),最终纳入年龄(OR=1.043)、机械通气时间(OR=1.140)、急性生理学与慢性健康状况评分表(APACHEⅡ)评分(OR=1.081)、血糖(OR=1.117)、乳酸(OR=1.459)、神经阻滞剂(OR=3.499)6个变量构建出风险预测模型。预测模型的公式为P=1/1+exp(-Z)=1/1+exp(8.808-0.042×年龄-1.252×神经阻滞剂的赋值-0.078×APACHEⅡ评分-0.110×血糖-0.378×血乳酸-0.131×机械通气时间),本预测模型ROC曲线下面积为0.896(95%CI 0.824~0.914),最大约登指数为0.577,灵敏度为0.754,特异度为0.823,截段值为0.503。模型验证结果:AUC=0.880,灵敏度为70.2%,特异度为88.1%,准确率为80.2%。结论本研究构建的ICU获得性衰弱风险预测模型具有较好的预测效果,可为临床筛选ICU获得性衰弱高危患者提供参考。

Objective To analyze the risk factors of Intensive Care Unit-Acquired Weakness,and to develop and verify the model.Methods A total of 247 patients admitted to ICU patients from November 2018 to October 2019 were selected,and risk factors between ICU acquired weakness group(n=106)and non-ICU acquired weakness group(n=141)were compared using logistic regression for model construction.The Hosmer-Lemeshow test was used to verify the goodness of fit of the model.The area under the ROC curve was used to test the model to predict the effects.From November 2019 to May 2020,106 patients were recruited for application of the model.Results The incidence of ICU acquired weakness in this study was 42.91%(106/247),and 44.34%(47/106),the study finally included age(OR=1.043),mechanical ventilation time(OR=1.140),APACHE II score(OR=1.081),blood sugar(OR=1.117),lactic acid(OR=1.459),and neuromuscular blockers(OR=3.499)to construct the risk prediction.The model formula was P=1/1+exp(-Z)=1/1+exp(8.808-0.042×age-1.252×neuromuscular blockers-0.078×APACHE II score-0.110×blood sugar-0.378×lactic acid-0.131×mechanical ventilation time.The area under the ROC curve of this model was 0.896(95%CI:0.824-0.914),the maximum value of the Youden index was 0.577,and the corresponding sensitivity was 0.754,the specificity was 0.823,the cutoff value was 0.503.The model verification results the sensibility of 70.2%,the specificity of 88.1%,and the accuracy of 80.2%.Conclusion The predictic model of ICU acquired weakness couducted in this study has satisfactory prediction effect,which can provide a reference for clinical screening of high-risk patients.