Traditional Chinese Medicine (TCM), with over 5,000 years of empirical practice, increasingly employs modern scientific frameworks such as randomized controlled trials (RCTs) to validate therapeutic claims, yet its research reliability hinges critically on robust statistical rigor. By systematically analyzing articles from Phytomedicine and Journal of Ethnopharmacology , this study evaluates statistical methodologies in TCM research, focusing on the adoption of advanced analytical techniques (e.g., multivariate modeling) versus reliance on basic methods (e.g., ANOVA) and identifies reporting gaps in trial design (e.g., sample size estimation). Key findings indicate that foundational statistical methods, such as one-way ANOVA, were predominantly used (83.4% of articles), whereas more advanced approaches appeared in only 34.6% of studies. However, methodological rigor should not be equated with statistical complexity. The selection of analytical techniques must be driven by the research objectives, data structure, study design, and the complexity of the scientific questions under investigation. Advanced methods are not inherently superior; rather, the most appropriate approach is the one that is methodologically justified and aligned with underlying assumptions. Notably, substantial deficiencies in trial design and reporting were observed. A striking 81.5% of studies lacked pre-specified power calculations or sample size justifications, raising concerns about statistical validity. Reporting transparency was similarly limited: 48.3% of articles did not adequately describe statistical procedures, and 69.8% failed to provide confidence intervals for primary effect estimates. Collectively, these limitations increase the risk of biased interpretation and undermine the robustness, reproducibility, and credibility of the findings. Strengthening statistical rigor—through improved trial design transparency and adoption of advanced methods—is essential to enhance the credibility of TCM research, mitigate biases, and foster its integration into evidence-based medicine, ultimately ensuring clinically meaningful and actionable therapeutic insights. 中医拥有超过五千年经验实践, 如今越来越多地采用随机对照试验等现代科学框架验证其疗效主张, 然而其研究可靠性仍高度依赖于统计严谨性. 本研究通过系统梳理《Phytomedicine》与《Journal of Ethnopharmacology》两本期刊的相关文献, 旨在评估中医药研究中统计方法学的应用现状.分析重点聚焦于多元建模等先进统计技术与方差分析等传统方法在应用上的差异, 并识别从试验设计阶段 (如样本量估算) 到数据分析过程中存在的 methodological 问题. 主要研究发现, 尽管以单因素方差分析为代表的基础统计方法被广泛应用 (见于83.4%的文献) , 但多种复杂统计方法联合使用的比例显著不足 (仅占34.6%) .试验设计层面存在普遍缺陷:高达81.5%的研究未进行先验功效分析或对样本量的合理性作出说明.同时, 统计报告的整体透明度偏低——近半数研究 (48.3%) 对统计流程的描述不够充分, 69.8%的研究未报告主要效应估计的置信区间.上述方法学上的不足可能导致对治疗效应的误判, 进而削弱研究结论的可信度. 提升试验设计的规范性与统计报告的透明度, 并推动先进统计分析方法的广泛应用, 是增强中医药研究可信度,降低偏倚风险的关键路径.这亦将促进中医药研究更好地融入循证医学体系, 最终为临床实践提供更具参考价值的治疗依据.
MacDonald et al. (Sun,) studied this question.