阶梯递进的辨证论治疗效评价方法研究

In this study, according to the characteristics of treatment based on syndrome differentiation such as basing on clinical practice, individualization, dynamics and complex intervention etc, by making full use of the data from clinical practice and discriminating evaluation elements such as intervention, appropriate population and evaluation indexes etc. gradually, the stepwise evaluation methods for determining the causal relationship between treatment based on syndrome differentiation and its effectiveness will be established.The study will take primary insomnia as a support and take clinical practical data from different famous veteran doctors in traditional Chinese medicine (TCM) as a starting point. The effectiveness evaluation course of treatment based on syndrome differentiation will be divided into three stages involving discovery, optimization and conformation of effective prescription of treatment based on syndrome differentiation. Retrospective exploratory study based on clinical practical data, prospective cohort study for optimization and randomized controlled trial for conformation will be applied correspondingly. To founding and identifying effective prescription of treatment based on syndrome differentiation, its appropriate population and evaluation indexes manifesting its effect advantage the methods including hierarchical topic model, gene set enrichment analysis, conditional mutual information and minimum reference set will be used. In order to increase evidence level of clinical study for primary insomnia treated by treatment based syndrome differentiation bias and confounding from effectiveness evaluation of treatment based on syndrome differentiation will be eliminated and controlled gradually through the progressive and stepwise design above to identify causal relationship between treatment based on syndrome differentiation and its effectiveness. The results will set a methodological demonstration for evaluating effectiveness of treatment based on syndrome differentiation scientifically and objectively and bring reliable scientific evidence for increasing recognition of TCM effectiveness.

本研究根据辨证论治以临床实践为基础，个体化、动态与复杂干预等特点，充分利用临床实际数据，通过逐步辨析干预措施、适宜人群、评价指标等评价要素，建立阶梯递进的确定辨证论治疗效因果关系的评价方法。研究以原发性失眠为载体，以不同名老中医辨证论治的临床实际数据为起点，将其辨证论治疗效评价过程分为辨证论治有效处方发现、优化及疗效验证三个阶段，依次采用基于临床数据的回顾性探索研究、前瞻性队列研究以及随机对照验证研究。应用分层主题模型、基因集富集度分析、条件互信息和最小参考集等方法，发现与确认辨证论治有效处方及其适宜人群和体现辨证论治疗效优势的评价指标。通过上述循序渐进、阶梯递进的研究设计，逐步消除和控制辨证论治疗效评价中的偏倚和混杂，以确认辨证论治疗效的因果关系，从而提高辨证论治失眠临床研究的证据级别。本研究结果将为科学客观评价辨证论治疗效提供方法学示范，为提高中医疗效公认度提供可靠的科学证据。

辨证论治是动态、个体化的复杂干预。目前常见的疗效评价研究是采用临床流行病学及设计、测量和评价方法，对固定干预措施在群体层次的效果进行评价，并利用“临床观察表”收集数据，建立系统的数据质量保障和控制体系。这是在“简单范式”下，抽样的、精确的、小数据的评价方法，实践证明其对于固定干预净效应的评价是有效的，结果是可重复的。对辨证论治中已经形成固定的专病专方的疗效评价是可行的，但对于因人、因时、因地辨证论治个体化、不固定方药的干预评价则往往是失效的。.本研究是借助大数据研究的思路，针对辨证论治个体化干预，采用真实世界临床数据采集与管理分析的方法，评价主导辨证论治的“医生”治疗某种疾病的疗效，评价结果推荐给人们的不再是一种固定的干预措施，而是某个个体化诊疗的“医生”。这是在“复杂范式”下，全样本的、混杂大数据的评价方法。不仅可以评价辨证论治的疗效，还可以发现个体医师诊疗的优势特点。由于“范式”的转化，收集数据的方法、质量控制和保障的体系、伦理保护等都将发生根本性的变化。