RESUMO
TCM clinical research covers different design types on basis of objective of the study. For exploratory research of TCM, cross-sectional study and descriptive research (e.g. case report and case analysis) is suitable. For hypothesis-testing research, RCT design is commonly adopted. RCT is most commonly used in activity or prevention therapy research, when using randomized cross-over design, the influence of sequence effects should be considered, and double-blind method is the best choice. Cohort study includes various intervention measures, as it doesn’t involve randomized allocation and blinding. Case-control study is widely applicable, but it may bring about bias. Cross-sectional study is much used in syndrome investigation of TCM clinical research. Descriptive research is available for summarization of TCM clinical experience and TCM exploratory research. Third party-central randomization system is recommended, as the randomization is not easy to be damaged. Randomization concealment should also be attached importance to. When blinding is impossible, blindness testing could reduce measurement bias, positive control should be supported high-level evidence. Placebo application could improve the level of evidence-based medicine.