Quick answer: What makes real-world evidence regulatory-grade?
Regulatory-grade real-world evidence (RWE) is evidence generated from real-world data (RWD) in a way that can withstand regulatory, HTA, or payer scrutiny. It is not created by using a large or reputable-looking dataset alone, and it is not a substitute for a randomized trial when a trial is feasible, nor a way to rescue a study that has already failed to meet its objectives. RWE requires four things: a documented fit-for-purpose real-world data source assessment, a protocol and statistical analysis plan finalized before the study starts, traceable and auditable data handling, and patient-level data availability where applicable.
If one of these areas is weak or undocumented, the evidence is more likely to be challenged during an FDA submission, HTA submission, or payer review, regardless of how large the underlying dataset is. Meeting these criteria demonstrates methodological rigor; it does not guarantee acceptance by any specific regulator, and different regulators, including the FDA and the EMA, do not always reach the same conclusion from the same data.
Key takeaways: Regulatory-grade real-world evidence
- “Regulatory-grade” is a specific, checkable, real-world evidence standard, not a marketing description.
- The four core quality signals are: fit-for-purpose real-world data source selection, a pre-specified protocol and analysis plan, traceable data handling, and patient-level data availability where applicable.
- The same regulatory-grade standard applies well beyond FDA submissions, HTA bodies and payers are effectively checking for similar real-world evidence quality signals in an HTA submission or value dossier, though meeting one regulator’s or body’s bar doesn’t guarantee another will reach the same conclusion from the same data.
- Medical device guidance finalized in December 2025 allows aggregate or de-identified real-world data in some cases; drug and biologic guidance has not adopted that flexibility yet.
- FDA medical device guidance finalized in December 2025 allows aggregate or de-identified real-world data in some cases; FDA drug and biologic guidance has not adopted that flexibility yet, and this is an FDA-specific position rather than one shared by the EMA or other regulators.
- A short written self-check against these four criteria, done before committing a budget to a real-world data source, catches most problems early.
How ‘regulatory-grade’ is defined for real-world evidence
“Regulatory-grade” is often used in real-world evidence conversations, sometimes as a defined standard and sometimes as a loosely applied marketing term. For a market access or evidence strategy team deciding how much to invest in a real-world data source, that ambiguity is a real problem. The difference between real-world evidence that holds up under FDA or HTA scrutiny and evidence that doesn’t comes down to four specific, checkable criteria, not the size or reputation of the underlying real-world data.
The four criteria for regulatory-grade real-world evidence
A regulatory-grade RWE study is built before the analysis begins. The data source, protocol, analysis plan, and data-handling process must be defensible before results are generated, so the final evidence can be reviewed as part of a planned study rather than as a post hoc justification.
Fit-for-purpose data source selection
Before a study starts, the data source itself needs a systematic, documented evaluation against the specific research question being asked. In practice, that means assessing the source’s data quality (how complete, accurate, and consistent the underlying data is), its relevance (whether it actually captures the population, exposure, and outcomes the research question requires), and its reliability (whether the same data would be captured consistently if the study were repeated).
FDA guidance is explicit that sponsors should justify why a given RWD source was selected or excluded, and be ready to discuss that justification before study implementation, not after results are in hand.
Pre-specified protocol and analysis plan
A statistical analysis plan finalized before the study runs, and a protocol registered in a public repository, both demonstrate that the study design wasn’t shaped by its own results. Because real-world data isn’t randomized, the analysis plan also needs to specify upfront how confounding will be identified and adjusted for, rather than leaving that decision until after the data has been seen. A retrospective analysis built after the fact carries a credibility problem that pre-registration avoids, however rigorous its underlying methodology.
Traceable, auditable data handling
Documented roles and responsibilities for every third party involved, and a maintained audit trail of the data itself, so the path from raw data to submitted conclusion can be reconstructed and checked by a reviewer who wasn’t part of the original study team.
This standard isn’t limited to structured RWD: Genesis applies the same traceability principle to literature-based evidence through EVID AI, where every extracted data point carries a citation back to its source, so a reviewer can check the same kind of audit trail behind a systematic review as behind a real-world data study.
Patient-level data availability, where applicable
For drug and biological product submissions, the FDA’s baseline expectation is that patient-level data can be made available for submission when needed. FDA medical device guidance finalized in December 2025 has started to relax this in specific cases, allowing aggregate or de-identified data, but that flexibility hasn’t yet extended to drugs and biologics, so it isn’t a safe assumption to build a drug submission strategy around yet, and it’s an FDA-specific position rather than one shared across regulators.
Even where the requirement to submit identifiable data is relaxing, regulators generally still expect an analysis that adjusts for individual-level confounders; HTA bodies have shown more openness to aggregate-level analyses, pooled published literature, for example, than regulators typically have.
Self-check: Is your real-world data source regulatory-grade?
Before a team commits budget to a real-world data source, it should be possible to answer four basic questions in writing. These questions will not replace a full feasibility or study design process, but they can quickly reveal whether the evidence strategy is being built on a defensible foundation.
| Question | Why It Matters |
|---|---|
| Has this data source been evaluated against this specific research question, with the reasoning documented? | Prevents a source from being chosen for convenience and justified after the fact. |
| Is the protocol and analysis plan finalized and registered before any data analysis begins? | Shows the study design wasn’t shaped by its own results. |
| Can every party that handled the data, and every step the data went through, be traced and audited? | Lets a reviewer reconstruct and check the path from raw data to conclusion. |
| Is patient-level data available for submission if the FDA or an HTA body asks for it? | Meets the current baseline expectation for drug and biologic submissions. |
If any row can’t be answered with a clear yes and a paper trail, that’s worth resolving before, not after, a budget is committed to the study.
Regulatory-grade real-world evidence in practice for HTA submissions and value dossiers
A regulatory-grade approach and a convenience-driven approach can start from the same raw data and end up in very different places. A team that picks a data source because it’s already licensed or familiar, runs the analysis, and writes the justification afterward to match what the data showed is building on the wrong foundation, even if the statistics are sound.
A team that documents why a specific data source fits the research question, registers the protocol before analysis begins, and can trace every step a reviewer might ask about is building evidence that holds up when someone outside the original team examines it. The FDA’s December 2025 device guidance illustrates the same principle: the 73 published examples of RWE-supported marketing authorizations from FY2020 to FY2025 share documented, traceable reasoning behind the data source choice, not just a large sample size.
Want a second opinion on whether your current real-world data source would meet this bar?
A weak fit-for-purpose assessment can compromise an otherwise well-designed RWE study. Genesis can help you review candidate real-world data sources, pressure-test whether they match your research question using structured approaches such as the Structured Process to Identify Fit-for-Purpose Data (SPIFD), and identify documentation gaps before a study is locked in.
Regulatory-grade RWE beyond FDA submissions: HTA and payer standards
The same four criteria apply well beyond a formal FDA submission. HTA bodies evaluating an HTA submission or value dossier, payers assessing a comparative effectiveness claim, and internal governance teams signing off on an evidence generation plan are all checking for the same underlying real-world evidence qualities: a documented rationale, a pre-specified design, traceable data handling, and defensible completeness.
A real-world data source and study design that clears the FDA’s fit-for-purpose bar is generally well positioned for an HTA submission or payer review too, though clearing one regulator’s or body’s bar is not a guarantee of acceptance by another. The FDA, EMA, and HTA bodies don’t always reach the same conclusion from the same data, which is why it’s worth building evidence generation plans around this regulatory-grade standard itself, rather than around any single regulator’s requirements..
How Genesis Research Group approaches regulatory-grade RWE
Genesis Research Group is a data-agnostic real-world evidence and HEOR consultancy, not tied to a single dataset, vendor, or methodology, and has worked across more than 40 unique real-world data sources for clients preparing FDA and HTA submissions. That range matters specifically for regulatory-grade work: a fit-for-purpose assessment only means something if the team doing it can genuinely choose the best-fit real-world data source from a wide field, rather than justifying whichever source they already have access to. It’s the same reasoning behind Genesis’s Flexible Integrated Team model: a team that stays with an asset across submissions can carry a fit-for-purpose assessment forward and adapt it, rather than starting the evaluation over with every new study.
FAQs: Regulatory-grade real-world evidence
These are the questions that market access and evidence strategy teams ask most often when they’re trying to determine whether a real-world data source or study design would withstand regulatory scrutiny.
What does ‘regulatory-grade’ real-world evidence actually mean?
Regulatory-grade refers to real-world evidence built to standards that hold up under regulatory or HTA scrutiny: a documented fit-for-purpose data source assessment, a pre-specified protocol and analysis plan, traceable and auditable data handling, and patient-level data availability where applicable.
Is regulatory-grade RWE only relevant for FDA submissions?
No. HTA bodies, payers, other regulators such as the EMA, and internal evidence governance teams are checking for similar underlying qualities, so the standard is worth applying to any evidence generation plan, not only formal regulatory submissions. That said, meeting the FDA’s bar doesn’t guarantee another regulator or body will reach the same conclusion from the same evidence.
Does meeting the FDA’s regulatory-grade criteria guarantee EMA or other regulator acceptance?
No. The FDA, EMA, and other regulators generally align on the same underlying principles, fit-for-purpose data, pre-specified design, and traceability, but they can and do reach different conclusions from the same real-world evidence. Regulatory-grade is a strong foundation, not a guarantee of acceptance by any specific regulator.
Can real-world evidence use aggregate or de-identified data and still be regulatory-grade?
For medical devices, FDA guidance finalized in December 2025 allows this in some cases, an FDA-specific position, not a broader regulatory consensus. For drug and biological products, the FDA’s 2023 guidance still expects patient-level data availability where applicable, so this flexibility hasn’t yet extended across the board, and most regulators still expect analysis that adjusts for individual-level confounders even as identifiable-data requirements relax.
Can real-world evidence use aggregate or de-identified data and still be regulatory-grade?
For medical devices, FDA guidance finalized in December 2025 allows this in some cases. For drug and biological products, the FDA’s 2023 guidance still expects patient-level data availability where applicable, so this flexibility hasn’t yet extended across the board.
What’s the difference between real-world data (RWD) and real-world evidence (RWE)?
Real-world data is the underlying data itself: electronic health records, claims data, registries, and similar sources. Real-world evidence is the clinical evidence about a product’s usage, benefits, or risks that is derived from analyzing that data. Not all RWD produces evidence that meets a regulatory-grade standard; the quality of the analysis and documentation determines that.
How long does a fit-for-purpose data source assessment take?
Data sources’ assessment times vary by therapeutic area and data source availability, but it’s a step worth properly resourcing rather than compressing, since a weak assessment here is the most common point of failure in an otherwise well-run RWE study.
Related reading: For the specific FDA guidance developments behind these regulatory-grade standards, see: How the FDA Actually Uses Real-World Evidence in Regulatory Decisions. For more on Genesis’s approach to real-world evidence generally, see Real-World Evidence Services, or browse the Genesis Insights hub.
Not sure if your current data source would hold up to regulatory scrutiny?
If you are planning an FDA submission, HTA submission, value dossier, or payer-facing evidence package, Genesis can help you assess whether your real-world data source, study design, and documentation are strong enough for the decision ahead.
Article sources
- FDA. Considerations for the Use of Real-World Data and Real-World Evidence To Support Regulatory Decision-Making for Drug and Biological Products. Final guidance, August 2023.
- Federal Register. Use of Real-World Evidence To Support Regulatory Decision-Making for Medical Devices. Final guidance, December 2025.
- FDA. M14 General Principles on Planning, Designing, Analyzing, and Reporting of Non-interventional Studies That Utilize Real-World Data for Safety Assessment of Medicines. March 2026.
- The National Library of Medicine. The Structured Process to Identify Fit-For-Purpose Data: A Data Feasibility Assessment Framework. Published December 2021.
