This paper makes an interpretation of the collection Acupuncture: how to improve the evidence base published by BMJ & BMJ Open. Studies show that the quality of randomized controlled trial (RCT) of acupuncture is low, and multivariable Meta-regression analysis fails to confirm most factors commonly believed to influence the effect of acupuncture. The methodological challenges in design and conduct of RCT in acupuncture were analyzed, and a consensus on how to design high-quality acupuncture RCT was developed. The number of acupuncture systematic reviews was huge but the evidence was underused in clinical practice and health policy, and a large number of western clinical practice guidelines recommended acupuncture therapy, but the usefulness of recommendations needed to be improved. In view of the problems in clinical research on acupuncture mentioned in this collection, combined with the analysis of the purpose of clinical research on acupuncture, perspectives, study types, as well as the relationship between evidence and clinical decision-making, a five-stage study paradigm of clinical research on acupuncture is proposed.
对《英国医学杂志》（BMJ）发表的针灸专辑“如何提升针灸证据质量”系列文章进行解读。研究显示针灸随机对照临床试验（RCT）报告质量较低，多因素Meta回归分析未能证实多数通常认为的针灸效应影响因素；分析了针灸RCT面临的方法学挑战，形成如何设计高质量针灸RCT方法学共识，系统评价证据数量巨大但未得到充分应用，大量西医临床实践指南推荐针灸疗法但推荐意见实用性有待提高。针对该专辑中提到的针灸临床研究存在的问题，结合分析针灸临床研究目的与研究视角、研究类型以及研究证据与临床决策的关系，提出了针灸临床研究的五阶段研究模式。.
Keywords: acupuncture; clinical practice guideline; paradigm of clinical research; randomized controlled trial (RCT); systematic review.