中南医学科学杂志2025，Vol.53Issue(3)：473-475,542,4.DOI:10.15972/j.cnki.43-1509/r.2025.03.023
脓毒症肾衰竭风险中医证素预测模型构建
Construction of traditional Chinese medicine syndrome element prediction model for sepsis-associated acute kidney injury risk
摘要
Abstract
Aim To construct a traditional Chinese medicine syndrome element prediction model for sepsis-associated acute kidney injury risk.Methods Clinical data of 189 sepsis patients were retrospectively analyzed.Patients were divided into acute kidney injury group(n=49)and non-acute kidney injury group(n=140)according to whether acute kidney injury occurred.Univa-riate analysis and Logistic regression analysis were used to screen traditional Chinese medicine syndrome elements.The prediction model was evaluated by Hosmer-Lemeshow test and ROC curve.Results In terms of traditional Chinese medicine disease location syndrome elements,the proportion of lung,spleen,and kidney involvement in the sepsis patients with renal failure was higher than that in the non-renal failure group(P＜0.05).Regarding traditional Chinese medicine disease nature syndrome elements,there were sig-nificant differences between the two groups in qi deficiency,yang deficiency,blood stasis,heat-fire,and water retention(P＜0.05).Multivariate Logistic regression analysis showed that spleen,kidney,qi deficiency,yang deficiency and water retention were independ-ent risk factors for sepsis-associated acute kidney injury(P＜0.05).The Hosmer-Lemeshow test showed good fit(P=0.578).The AUC of ROC curve was 0.865 with 90.14％sensitivity and 85.87％specificity.Conclusion The constructed traditional Chinese medicine syndrome element prediction model shows good predictive performance for sepsis-associated AKI risk.关键词
脓毒症/肾衰竭/中医证素/预测模型/ROC曲线Key words
sepsis/acute kidney injury/tradtional Chinese medicine syndrome elements/prediction model/ROC curve分类
医药卫生引用本文复制引用
申建国,王健,陈羽..脓毒症肾衰竭风险中医证素预测模型构建[J].中南医学科学杂志,2025,53(3):473-475,542,4.基金项目
河北省中医药管理局科研计划项目(2022232) （2022232）
