Objective To review the current application of sample size estimation in real-world studies (RWS), analyse parameter settings and commonly used methods, and provide methodological guidance for researchers conducting RWS. Methods First, ClinicalTrials.gov was searched to identify RWS with documented sample size calculations. Key information was extracted for descriptive analysis. Secondly, critical parameters and common estimation methods for RWS sample size calculations were systematically reviewed, and strategies were proposed for addressing common challenges. Finally, relevant international reporting standards were interpreted. Results The literature review included 44 clinical trials with a wide range of sample sizes (30 to 30 400 cases). While most studies detailed the sample size estimation process, the parameter settings were often incomplete and many failed to adequately consider the characteristics of real-world data. Therefore, we proposed key parameters for RWS sample size estimation, including effect size, significance level and statistical power. Researchers should also consider issues such as heterogeneity, confounding factors and data quality. This study clarified the essential elements of reporting sample size estimation. Conclusion Methodological guidance for real-world evidence sample size estimation is lacking. We advise researchers to standardise reporting procedures for sample size estimation in future studies and to set parameters reasonably based on research objectives, study design types and data characteristics. This will enhance the transparency and scientific rigour of real-world evidence.
Structured template and reporting tool for real world evidence (STaRT-RWE) was developed by a team led by professor Shirley V Wang of Brigham and Women's Hospital, Harvard Medical School, which is to plan and report on the implementation of real world evidence (RWE) studies on the safety and efficacy of treatments. The template, published in the journal BMJ in January 2021, has been endorsed by the International Society of PharmacoEpidemiology and the Transparency Initiative promoted by the International Society of Pharmacoeconomics and Outcome Research. This article interprets its entries to promote the understanding and application of STaRT-RWE by domestic scholars engaged in real world study, and help to improve the transparency, repeatability, and accuracy of RWE research.
Assessment of Real World Observational Studies (ArRoWS) is a tool developed by the Leicester Real World Evidence (LRWE) Unit of the Diabetes Research Centre of the University of Leicester in the United Kingdom to assess the quality of real world evidence research, and has been reported to have good practicability. ArRoWS can be used to quickly and specifically assess the quality of real world evidence research that uses electronic health record information. The tool contains 16 items, nine of which are common items, and seven of which are related to specific research designs. The current study introduces the development background, development process, assessment items, assessment criteria, and application methods of ArRoWS and other related aspects, to provide references for real world researchers in China.
The active comparator, new user (ACNU) design is an important design developed under the concept of the target simulation experimental framework. It aims to reduce indication confounding, immortal time bias, prevalence-incidence bias, and other unmeasured confounders by simulating head-to-head randomized controlled trials. It is widely applied in scenarios such as comparing the efficacy of newly marketed drugs with existing standard treatments, evaluating drug safety and adherence, exploring drug repurposing, and optimizing algorithms for processing medical big data. This article introduces the application and practice of the ACNU design in real-world data research from aspects such as concept, development, advantages and disadvantages, and implementation points, and also presents an outlook on its application in the field of traditional Chinese medicine. It is believed that with the progress in understanding the design of observational studies of real-world data, the ACNU design is expected to be more widely applied and provide new ideas for researchers' scientific research designs.