首页|基于隐结构模型的卒中相关重症肺炎证素及辨识标准研究

基于隐结构模型的卒中相关重症肺炎证素及辨识标准研究

扫码查看
目的 基于隐结构模型探讨卒中相关重症肺炎的证候要素,初步建立卒中相关重症肺炎的中医辨识标准,为临床辨证提供参考依据.方法 收集2017 年1 月至2021 年12 月间于天津中医药大学第一附属医院南院重症监护室住院治疗的卒中相关重症肺炎患者 230 例,依据出现频率(>10%)纳入相关的四诊信息及症状体征.运用隐结构分析软件Lantern5.0 对纳入的 230 例卒中相关重症肺炎证候数据进行双步隐树模型(BI算法)分析,获取隐结构模型.通过显变量和隐变量相关的结构图、两两互信息曲线和隐类概率分布情况,运用中医基础理论对隐类进行诠释后再进行综合聚类分析,对每个证候的临床症状和体征赋予分值,并确立一个阈值,建立打分形式的辨证规则.结果 (1)共纳入78 个相关的临床特征信息,通过隐树模型分析后,得出21 个隐变量,46个隐类,初步得到卒中相关重症肺炎的病性类证候要素为:痰证、热证、湿证、气虚证、阴虚证、阳虚证.(2)通过综合聚类分析,得到卒中相关重症肺炎的辨识标准如下:痰证(阈值:26);热证(阈值:21);湿证(阈值:38);气虚证(阈值:6.3);阴虚证(阈值:46.6);阳虚证(阈值:15.4).结论 (1)卒中相关重症肺炎患者的常见病性类证候要素:痰证、热证、湿证、气虚证、阴虚证、阳虚证;(2)初步建立了卒中相关重症肺炎的常见病性类证候要素的辨识标准.
The study of the common syndrome elements and identification criteria of stroke-associated severe pneumonia based on Latent Class Analysis
Objective To explore the syndrome elements of stroke-associated severe pneumonia,establish the identification standard of stroke-related severe pneumonia,provide objective basis for accurate clinical syndrome differentiation treatment.Methods 230 patients with stroke-related severe pneumonia hospitalized in the South Intensive Care Unit of the First Teaching Hospital of Tianjin University of Traditional Chinese Medicine(TCM)were collected from January 2017 to December 2021,and the in-formation of four methods of examinations in TCM and some physical signs of 230 patients were extracted from the electronic medical records.Lantern5.0 was used to analyze the data according to occurrence frequency(>10%)of symptoms and signs and the relevant structure model was established.The structure diagram,mutual information curve and latent class probability distribution table of latent variables and explicit variables were analyzed.The results were interpreted using TCM basic theories.The common syndrome elements of stroke-associated severe pneumonia were obtained.Lanten5.0 software was employed to analyze the clinical features of each syndrome elements and the identification criteria with a threshold value were determined.Results A total of 78 relevant clinical manifestations were performed.And 21 latent variables and 46 subtypes were obtained.The disease syndrome elements of stroke-associated severe pneumonia were preliminarily obtained as follows:phlegm syndrome,heat syndrome,wetness syndrome,Qi deficiency syndrome,Yin deficiency syndrome and Yang deficiency syndrome.Through comprehensive cluster analysis,the preliminary identification criteria of stroke-associated severe pneumonia with a threshold were established.Conclusions The latent structure method was used to analyze the symptoms and signs of 230 cases of stroke-associated severe pneumonia,and the common syndrome elements of stroke-associated severe pneumonia were extracted as follows:phlegm syndrome,heat syndrome,wetness syndrome,Qi deficiency syndrome,Yin deficiency syndrome and Yang deficiency syndrome.Through the method of comprehensive cluster analysis,the identification standard of disease syndrome elements of stroke-associated severe pneumonia was preliminarily established to provide objective basis for clinical syndrome differentiation.

stroke-associated severe pneumoniaLatent Class Analysissyndrome elementsidentification criteria

牛丽青、岳琳凯、肖璐、刘新桥

展开 >

300381 天津中医药大学第一附属医院急症部 国家中医针灸临床医学研究中心

卒中相关重症肺炎 隐结构法 证候要素 辨识标准

天津中医药大学第一附属医院"拓新工程"基金

201928

2024

环球中医药
中华国际医学交流基金会

环球中医药

CSTPCD
影响因子:1.553
ISSN:1674-1749
年,卷(期):2024.17(1)
  • 8