预测艰难梭菌感染的列线图模型构建

The construction of nomogram model for predicting Clostridioides difficile infections

目的分析腹泻患者中艰难梭菌感染(CDI)的危险因素,构建预测CDI的列线图模型。方法回顾性分析2015年8月至2019年7月1055例腹泻患者的临床资料,其中181例确诊为CDI。比较CDI患者与非CDI患者的粪便常规,血常规和血生化的差异。采用多因素Logistic回归分析筛选CDI的危险因素,并基于筛选出的危险因素建立腹泻患者中CDI发生风险的列线图预测模型,通过受试者操作特征曲线评估模型性能。结果1055例腹泻患者中CDI发生率为17.2%。Logistic回归分析结果显示中性粒细胞百分比、总蛋白和钾离子浓度是腹泻患者中提示CDI的独立危险因素(P<0.05);将独立危险因素引入R软件(R3.6.3)构建列线图模型,曲线下面积为0.717。结论基于腹泻患者中CDI的危险因素如中性粒细胞百分比、总蛋白和钾离子浓度建立的列线图预测模型具有良好的区分度,可为CDI的初筛提供指导价值。

Objective To analyze the risk factors of Clostridioides difficile infection(CDI)in patients with diarrhea and construct a nomogram model for predicting CDI.Methods A retrospective analysis of 1,055 patients with diarrhea in Hangzhou First People's Hospital affiliated to Zhejiang University School of Medicine from August 2015 to July 2019,among which 181 cases were diagnosed with CDI.We compared the differences in stool routine,blood routine and blood biochemistry between patients with CDI and non-CDI.We sampled multivariate logistic regression analysis to screen the risk factors of CDI,and established a nomogram prediction model of the risk of CDI in patients with diarrhea.The model performance was evaluated through receiver operating characteristic(ROC).Results The incidence of CDI in 1055 patients with diarrhea was 17.2%.Logistic regression analysis showed that the percentage of neutrophils,total protein and potassium were independent risk factors for CDI in patients with diarrhea(P<0.05).The independent risk factors were introduced into the R software(R3.6.3)to construct nomogram model,the area under the curve is 0.717.Conclusion The nomogram prediction model based on the risk factors of CDI in patients with diarrhea,such as the percentage of neutrophils,total protein and potassium,has a good degree of discrimination and can provide guidance value for the initial screening of CDI.