Home  /  News & Events  /  Events

7th Reproducibility Rounds webinar: Non-Reproducibility of Real-World Data Research: A Case Study

Real-world data platforms are increasingly used at research institutions due to their accessibility and rich, contemporaneous data across healthcare systems. These platforms lower barriers to research and are especially attractive to early-stage investigators who might have limited data gathering resources. However, real-world data studies pose significant design and analytic challenges that must be addressed to ensure rigor and reproducibility. In this talk, we use TriNetX as case study of a real-world data platform given its ease of access to data and analytic tools at many academic health centers. We investigate the composition of the research team and intent (causal vs. non-causal intent) of several hundred papers published in 2024. We select a single paper with no quantitative experts in the author list to investigate various aspects of methods, results, and inferential reproducibility. We end by posing questions to the audience to generate discussion on strategies to improve rigor and reproducibility with self-service tools for real-world data in medicine.