Aiming at the problems of uneven reporting quality and lack of unified standards in SMART in the context of personalized medicine, this paper analyzes the application and reporting status of SMART, and summarizes the key problems of current research reporting. These include unclear description of core design elements, missing information on randomization and blinding, insufficient reporting of phase transition rules, opaque sample size estimation and multiple comparisons, and lack of detailed description of data processing procedures. Based on the CONSORT standard, this paper proposed to expand the reporting specifications for SMART design methodology. By adding and elaborating items, the standardized description of multi-stage randomization process, decision nodes, embedded dynamic treatment regimes comparison framework and dedicated flow chart are emphasized to enhance the transparency, reproducibility and reliability of the results of SMART study. This study aims to provide structured and standardized reporting recommendations for SMART trials, and promote the output of high-quality evidence and clinical transformation of dynamic treatment strategies.
The sequential multiple assignment randomized trial (SMART) is an individualized medical model that adjusts intervention strategies in real time based on individual characteristics and stage-specific efficacy, applying to dynamic treatment regimens (DTRs) of "treatment based on syndrome differentiation" in traditional Chinese medicine (TCM). Focusing on two-stage asymmetric SMART designs with continuous outcomes, this paper reviews sample size estimation methods for four research objectives (comparing first-stage main effects, comparing second-stage main effects, comparing two DTRs, and selecting the optimal DTR), and details calculation formulas and parameters. Integrating the characteristics of TCM clinical research, this paper analyzes methodological considerations regarding parameter setting, dynamic decision logic, and tailoring variable selection within the context of syndrome differentiation. Finally, a TCM insomnia treatment example demonstrates the application. This study aims to provide a methodological reference for conducting SMART designs in TCM research, guiding researchers to estimate sample sizes accurately.