Comment on Framework for FDA's Real-World Evidence Program - part 1 (ENG)
The FDA’s perspective on real-world evidence (RWE) appears quite cautious, possibly because the guideline was initially written in December 2018. Real-world data refers to safety and effectiveness data collected outside traditional clinical trials (explanatory trial).Electronic health records (EHR) and insurance claims data can also be considered as real-world data.
It was notable that pragmatic trials are considered as RWE. Pragmatic trials are clinical trials conducted in settings similar to routine clinical practice. PRESIS-2 scores trials based on how closely they align with either explanatory (ideal environment) or pragmatic (real clinical routine) approaches. Rather than viewing this as a dichotomy, it’s seen as a continuum. The framework also mentions hybrid designs that combine traditional clinical trials with real clinical practice approaches (using medical claims/EHR for data collection).
Real-world evidence is used to improve clinical trial efficiency, including hypothesis generation for randomized controlled trials, determining prior probability distributions, and identifying target populations where interventions work best. Historically, RWE has been used to verify drug safety. While its use for determining effectiveness was initially limited, RWE usage is now expanding.
Regarding effectiveness, there are cases where single-arm intervention trials with external control arms were approved when parallel control groups were either unethical or impossible due to low prevalence and severity of disease. Blincyto received accelerated approval based on single-arm trial results using historical data. There are cases where pragmatic trial elements were used to generate RWE.
The framework emphasizes that efficacy and safety shown in clinical trials must translate into effectiveness and safety in the real world, though this replication isn’t always achieved. There’s hope that advances in analytical methodology and data management will improve the reproduction of clinical trial results in observational studies.
Article written by Presenter, SooYoung Kwak
Source: Original Korean article | English translation by Claude 3.5 Sonnet (New)