Page 73 - 2023-03-中国全科医学
P. 73
·322· http: //www.chinagp.net E-mail: zgqkyx@chinagp.net.cn Jaunary 2023, Vol.26 No.3
hierarchical cluster analysis,and the potential syndromes were inferred from the results. Using Lantern 5.0 software,based on
the LTM-EAST(Latent Tree Model-Extension Adjustment Simplification until Termination algorithm),a latent structure model
was constructed and a comprehensive cluster analysis was performed to infer potential syndromes. Then the common syndromes
and symptom characteristics of CVA were obtained by combining the results of frequency statistics,hierarchical clustering,
and latent structure analysis. Results A total of 621 studies were included,involving 115 symptoms,including cough,
pharyngeal itching,low sputum volume,etc. Thirty-nine syndromes were involved,among which the highly frequent syndromes
were pathogenic wind invading the lung,wind and cold invading the lung,and lung yin deficiency. Six potential syndromes
were derived from hierarchical cluster analysis,including wind and heat invading the lung,liver fire invading the lung,
lung qi deficiency and so on. The latent structure analysis was performed on the symptoms with frequency over 3%,by which
21 latent variables,and 42 latent classes were obtained,and 12 potential syndromes were further inferred by comprehensive
clustering,including pathogenic wind invading the lung,wind and cold invading the lung,wind and heat invading the lung,
and so on. Based on the above analyses,11 common syndromes were finally identified. Each syndrome has specific symptom
characteristics,for example,the syndrome of pathogenic wind invading the lung manifests itself as cough,sneezing,running
nose,and nasal congestion. Conclusion We identified 11 common syndromes of CVA with typical symptom characteristics,
which are pathogenic wind invading the lung,wind and cold invading the lung,wind and heat invading the lung,wind and
dryness damaging the lung,phlegm and heat congesting in the lung,phlegm and dampness accumulating in the lung,liver fire
invading the lung,cold fluid retention in the lung,lung qi deficiency,lung yin deficiency,and lung-spleen qi deficiency. Our
analysis will provide a basis for clinical syndrome differentiation and clinical investigation of CVA.
【Key words】 Asthma;Cought variant asthma;Respiratory tract diseases;Symptom complex;Latent structure
models;Cluster analysis
咳嗽变异性哮喘(cough variant asthma,CVA)是 研究、医家经验、个案报道;(2)研究对象为 CVA 患
一种以咳嗽为主要或唯一症状的特殊类型哮喘,我国 者;(3)年龄≥ 16 岁;(4)涉及证候或症状分布。
近 1/3 的慢性咳嗽与之相关 [1-3] ,30.0%~35.7% 的 CVA 1.2.2 排除标准 (1)无法获取全文者;(2)重复发
患者会在 5 年内演变成典型哮喘 [4-5] 。中医药在治疗 表的文献(仅保留 1 篇)。
CVA 方面疗效确切,能够有效改善患者临床症状、提 1.3 数据库的建立与规范 应用 SPSS 25.0 软件,提取
高生活质量、减少 CVA 急性发作次数等 [6-11] 。辨证论 合格文献的标题、年份、期刊、作者、证候、症状等信
治是中医立法处方、诊疗疾病的关键环节 [12] ,开展证 息以建立数据库。参照《中医临床诊疗术语第 2 部分:
候诊断标准研究对于证候辨识度及中医整体辨治水平的 证候(修订版)》 [15] 和《实用中医诊断学(精)》 [16]
提高具有重要意义 [13] 。临床常见证候的选择是证候诊 对证候名称进行规范,病因病机相同的证候统一证候名,
断标准研究中的关键问题之一,其直接影响着所形成标 如“风寒犯肺证”“风寒束肺证”“风寒袭肺证”统一
准的准确性、实用性以及是否易于推广等 [14] 。基于此, 为“风寒犯肺证”;参考《中医药学名词 -2004》 [17]
本研究通过检索相关文献,初步探究 CVA 常见证候及 对症状内涵相同或差异较小的名称进行规范和统一,如
其症状分布规律,以期为 CVA 临床实践及后续中医证 “痰黄”“痰色黄”“痰深黄”统一为“痰黄”。
候诊断标准建立相关的临床调查研究提供参考。 1.4 统计学方法 采用 SPSS 25.0 软件对症状、证候进
1 资料与方法 行频次与频率描述,其中频率为频次与条目数之比。
1.1 文献检索 计算机检索中国知网(CNKI)、万方 选择频率 >5% 的症状进行因子分析,基于最大方差法
数据知识服务平台、维普网(VIP)、中国生物医学文 得到旋转因子载荷矩阵,提取特征值 >1 的公因子作进
献服务系统(SinoMed)收录的涉及 CVA 证候或症状的 一步系统聚类分析,根据聚类结果推断潜在证候。采用
相关文献。检索式以 CNKI 为例:(主题词:咳嗽变异 孔明灯 Lantern 5.0 软件 [18] ,基于双步隐树分析 LTM-
性哮喘 OR 咳嗽变异型哮喘 OR 咳嗽性哮喘 OR CVA OR EAST 算法,选择频率 >3% 的症状构建隐结构模型,采
cough variant asthma OR 风哮)AND(主题词:中医 OR 用直连分析进行模型诠释,并绘制信息曲线图;结合医
证候 OR 证型 OR 症状)NOT(主题词:小儿 OR 儿童), 学专业知识及研究小组讨论,对隐变量进行综合聚类,
匹配选择“精确”,其余各项选择默认设置。检索日期 推断潜在证候,采用泛连分析对综合聚类模型进行诠释。
为建库至 2021-10-02。 应用模型贝叶斯信息准则(BIC)评分 [19] 评价所得模型,
1.2 纳入与排除标准 选择负分计量,分值越大提示所得模型越好。
1.2.1 纳入标准 (1)文献类型为临床观察及干预性 1.5 常见证候及其症状的确定