Feasibility of Using Real-World Data to Replicate Clinical Trial Evidence

This study found that only 15% of the US-based clinical trials published in high-impact journals in 2017 could be feasibly replicated through analysis of administrative claims or EHR data. This finding suggests the potential for real-world evidence to complement clinical trials, both by examining the concordance between randomized experiments and observational studies and by comparing the generalizability of the trial population with the real-world population of interest.