Incorporating Traditional Chinese Medicine Syndrome Differentiation in Randomized Trials: Methodological Issues (Cochrane {CAM} Field Invited Commentary)

摘要

Purpose: In traditional Chinese medicine (TCM) practice,decision on prescription is based on a process called BianZheng Lun Zhi (syndrome differentiation guided treatmentdecision). The syndrome differentiation process may notbe recognized in conventional standards of randomizedcontrolled trial (RCT), limiting the model validity and generalizabilityof results.Methods: This article discussed how syndrome differentiation,a classical TCM approach in diagnosis, can beincorporated into RCT design.Results: Four methodological solutions were proposed: (i)Lesson learnt from the design of patient reported outcomequestionnaire can inform how TCM diagnosis instrumentcould be developed. A proper TCM diagnostic tool with soundpsychometric properties can reduce variation in the syndromedifferentiation process. (ii) Treatment strategies for a specificTCM diagnosis could be highly diversified. Delphi techniquecan inform the design of optimal treatment program by facilitatingconsensus among experts. (iii) Subgroup analysis isoften needed in RCT recruiting patient with several TCMdiagnosis. It is highlighted that investigators should considerwhether the design, analysis and context of the trial are robustenough to support a reliable claim of subgroup effect associatedwith a particular TCM diagnosis. (iv) Finally,we discussedalternative research and analysis approaches for handlingmisalignment of Western and TCM diagnoses, including thepossibility of unifying TCM syndrome with Western phenotypesusing latent class analysis.Conclusion: Further methodological advances are neededin the better alignment of classical TCM theories and diagnosticinstrument development, aswell as in reducing bias duringthe expert consensus processes.