基于数据挖掘技术和Delphi法的小儿肺炎郁热辨证研究

Hypothesis Evidence: Childhood Pneumonia is a common and frequently-occurring disease, which is classified into the field of Warm Disease in Chinese Medicine. Chinese Medicine (TCM) has good result in it, further researchment should be encouraged. Our team carried long-term researchment on Childhood Pneumonia (CP) and Study of Warm Disease. As the previous Scientific Funding Project of CP to be an example, we innovatively used technique Data Mining for the Study of TCM Syndrome and Mechanism together the clinical treatment evaluation, discovered Stagnation, Heat and Phlegm are the basic mechanism of the illness. To carry on, we suggested theory like Heat Stagnated Syndrome Differentiation (SyD) is suitable for CP, for the deeper researchment. Method: Collect multi-centered and large number of clinical CP samples, with Heat Stagnated SyD and Data Mining together with Delphi Method, retrospective and prospective researchment of Heat Stagnated SyD in CP can be carried. Basis and Feasibility: Our team completed 1 project form National Natural Science Foundation of China, 3 Standardizations Editions for TCM industry, 9 projects and 40 more medical essays by the application of Data Mining and Delphi Method, also based on the TCM Mechanism Study and standardizations. Innovation: For the first time the use of Data Mining technique and Delphi Method on research of CP with Heat Stagnated SyD. Significance: Mathematical Research of CP could be more systematic and completed, which is benefit for Chinese Medicine researches to be more standardized and objective. Meanwhile, the level of prevention and treatment of Childhood Pneumonia in local area could be raised, so to enhance and protect the economy development.

立项依据：小儿肺炎为常见病和多发病，属中医温病学范畴。中医药治疗效果良好，须加强研究。本课题组长期开展小儿肺炎和温病学研究，在前一科学基金课题中以小儿肺炎为例，创新运用数据挖掘技术研究中医证候病机学和临床疗效评价基础，发现郁、热、痰为该病基本病机，进而提出小儿肺炎郁热辨证的假说，拟延续研究。研究内容：收集大样本、多中心小儿肺炎临床数据，运用郁热辨证与数据挖掘和Delphi法相结合，进行回顾性和前瞻性的小儿肺炎郁热辨证研究。基础及可行性：通过数据挖掘和Delphi法在中医病机学、标准化中的应用，已完成1项国家自然科学基金课题和3项中医药行业标准。完成课题9项，发表论文40余篇。创新点：首次运用数据挖掘技术和Delphi法开展小儿肺炎郁热辨证研究。意义：本课题将使小儿肺炎的数理研究更加全面和系统，并促进中医学客观化和规范化研究。同时，提高本地区小儿肺炎防治与研究水平，保障和促进社会经济发展。

项目背景：本课题组长期开展小儿肺炎和温病学研究，在前一科学基金课题中以小儿肺炎为例，创新运用数据挖掘技术研究中医证候病机学和临床疗效评价基础，发现郁、热、痰为小儿肺炎的基本病机，进而提出小儿肺炎郁热辨证的假说，拟延续研究。.主要研究内容：基于中医基础理论和温病学郁热病机理论，收集大样本、多中心小儿肺炎临床资料，设计郁热证候调查表，应用数据挖掘技术，对原有3250例临床资料做回顾性研究；新收集1500例做前瞻性研究，并将两组数据分析结果综合，开展Delphi法研究，探讨小儿肺炎的郁热辨证内容。.重要结果：通过聚类分析，发现小儿肺炎的类别与郁热辨证的分型有高度相似。通过关联规则分析，认识到各个郁热证型中权重占比较大的症状体征，以及高概率共现的症状体征，在证候表述中应增加2项或3项的症状体征关联项的表述。通过Delphi法研究，专家认为郁热辨证方法适用于小儿肺炎的临床辨证，可以扩展到儿科其他疾病及温病学的应用。综合以上研究，认识到小儿肺炎郁热辨证内容及温病学郁热辨证研究的思路与方法。.关键数据：关联规则运算得到超百万条规则，获得大量2～10项关联的频繁项集。运用Delphi专家评估统计，发现卫分无形郁热证，气分无形郁热证，营分无形郁热证，痰热郁结证，湿热郁结证，热瘀互结证的 值介于1.45~1.85之间，S值依次为31，35，32，37，29， 33。R%值介于0%~15%之间。积滞热结证和燥屎热结证的S<=26，同时R%为25%、30%。.科学意义：一是郁热辨证对温热病辨证思路与方法具有广泛的启发作用，可以扩大应用到小儿温病及温病学范畴，若能深入和拓展研究将是温病学理论的重大创新；二是应用数据挖掘方法研究中医辨证，在证候表述中增加2项或3项的症状体征关联项的表述，符合中医证候学特点和规律，对于中医证候学研究具有创新性和实用价值；三是本课题研究方法对于中医辨证方法的研究具有启发和示范作用。