针灸研究中组间症状积分的定量比较与组内分级定性法利弊的探讨

Adequate application of quantitative and qualitative statistic analytic methods in acupuncture clinical trials

近年来,在中医针灸随机对照临床试验(randomized controlled trial,RCT)中恰当地应用统计学方法日益受到重视。基于假设检验的统计推断是临床试验与循证医学的基础,根据临床试验设计的基本统计原则,本文探讨了中西方文献中针灸RCT设计与分析方法学的差异。国内针刺临床试验多采用3级或4级定性法进行临床统计分析,而国外的临床试验则多采用组间症状积分的定量比较。本文以针刺治疗疼痛的临床试验为例,讨论了两种分析方法的利弊。我们认为两种方法可以兼用,但应以定量分析方法为主,分级定性法为辅,以取各自方法的所长。本文的目的是希望通过提出问题来抛砖引玉,以期引起更多的学术讨论,从而提高中医针灸临床研究的科学性和严谨性。

Recently, proper use of the statistical methods in traditional Chinese medicine （TCM） randomized controlled trials （RCTs） has received increased attention. Statistical inference based on hypothesis testing is the foundation of clinical trials and evidence-based medicine. In this article, the authors described the methodological differences between literature published in Chinese and Western journals in the design and analysis of acupuncture RCTs and the application of basic statistical principles. In China, qualitative analysis method has been widely used in acupuncture and TCM clinical trials, while the between-group quantitative analysis methods on clinical symptom scores are commonly used in the West. The evidence for and against these analytical differences were discussed based on the data of RCTs assessing acupuncture for pain relief. The authors concluded that although both methods have their unique advantages, quantitative analysis should be used as the primary analysis while qualitative analysis can be a secondary criterion for analysis. The purpose of this paper is to inspire further discussion of such special issues in clinical research design and thus contribute to the increased scientific rigor of TCM research.