中医临床试验证据个体适用性评价方法的建立

Whether the evidence from clinical trial is applicable for individual patients is the key point of evidence-based practice, and clinical efficacy is the ultimate standard for judging evidence applicability. Taking a look into the complexities involved in the use of evidence from traditional Chinese medicine (TCM) clinical trials, the present study aims to establish a method for evaluating the applicability of TCM research evidence in individual patients by taking prediction of efficacy as pointcut, based on treatment group raw data from a clinical trial of Maziren pills on functional constipation (excess heat pattern). Firstly, considering the clinical practical significance, principal component analysis model, systematic review, expert consultation and patient scoring are used jointly to generate the comprehensive weights of the evaluation index for patients in the treatment group. Then the weights are introduced into the grey relational analysis model to assess and rank efficacy in individual patients. 3 efficacy hierarchies which are significant efficacy, general efficacy and no efficacy will be determined according to Delphi consensus. Secondly, the weights of characteristics index will be determined by the analytic hierarchy process, and the grey relational analysis model will be used again for evaluating the similarity between a new individual patient and each efficacy hierarchy. The goal is to predict clinical efficacy of new individual patient and determine evidence applicability. Finally, a case series study will be carried out to verify the reliability of the evaluation method. This study will provide a methodological demonstration for the accurate and rational application of evidences from clinical trials of TCM.

临床试验证据能否适用于患者个体是循证实践关注的要点，而临床疗效是判断试验证据个体适用性的最终标准。本项目以疗效预测为切入点，着眼于中医临床试验证据应用的复杂性因素，依托于麻子仁丸治疗实热型功能性便秘临床试验的试验组原始数据，构建中医临床试验证据个体适用性评价方法。首先从临床实际意义出发，联合运用主成分分析模型、系统评价、专家咨询和患者评分法，获得试验组患者个体化评价指标的综合权重，将权重加入灰色关联分析模型，评价试验组患者个体疗效并排序，再以德尔菲法将个体疗效序列划分为显效、有效和无效的疗效阶层；继而以层次分析法确定特征指标的权重，再次将权重加入灰色关联分析模型，计算新患者与各疗效阶层的特征接近程度，预测新患者疗效倾向，评价证据的个体适用性；最后以病例系列研究验证评价方法的可靠性。这将为中医临床试验证据的准确合理应用提供方法学示范。

临床试验证据能否适用于患者个体是循证实践关注的要点，而临床疗效是判断试验证据个体适用性的最终标准。本项目分析在中医药领域开展临床试验证据个体适用性评价研究的理论基础，以疗效预测为切入点，依托中医药临床试验的原始数据，引入各类数学模型工具，示范个体化疗效预测流程，建立并验证中医临床试验证据个体适用性评价方法，以期能规范地辅助中医临床的循证诊疗决策。项目在执行中完成了理论分析、数据预处理、专家权重调查、模型构建、应用验证各环节的方法学示范，并扩展探索了支撑疗效预测的多种模型算法，建立了“临床试验证据决策辅助系统”的网页端、手机端工具，面向未来的推广应用，还兼顾探讨了基础研究数据、数据缺失和样本偏小等因素对应用本项目成果的价值或影响。项目为循证临床决策框架下的中医规范辅助诊疗提供了一种策略。