FDA's New RWE Guidance: How De-Identified Data is Reshaping Submissions

In a landmark move that signals a new era for evidence-based regulatory decision-making, the FDA has released guidance that significantly lowers the barriers to using Real-World Evidence (RWE) in drug and medical device submissions. The key change: sponsors will no longer need to provide individual patient-level identifiable data in many RWE applications. This article breaks down what the guidance says, why it matters, and what sponsors need to do now.

Background: The Promise and Challenges of RWE

Real-World Evidence—derived from sources like electronic health records (EHRs), insurance claims, patient registries, and wearable devices—has long been seen as a game-changer for drug development. Unlike data from randomized controlled trials (RCTs), which are conducted under tightly controlled conditions, RWE reflects how treatments work in diverse, real-world patient populations.

The FDA has been encouraging the use of RWE for years. The 21st Century Cures Act (2016) mandated that the agency develop a framework for evaluating RWE in regulatory submissions. Since then, RWE has been successfully used to support new indications, post-market studies, and even some initial approvals.

However, a significant barrier remained: the FDA's historical requirement for access to individual patient-level data (IPD) in many RWE submissions. This created challenges related to patient privacy, data sharing agreements, and the logistical complexity of transferring large datasets to the agency.

What the Guidance Says

The guidance, titled "Real-World Data: Use of De-Identified and Aggregate Data in Regulatory Submissions," provides clarity on several key points:

1. When De-Identified Data is Acceptable

The FDA will accept de-identified data for RWE submissions when:

• The de-identification process follows accepted standards (e.g., HIPAA Safe Harbor or Expert Determination methods).
• The analytical methods used are transparent and reproducible.
• Sufficient documentation is provided on data sources, data quality, and limitations.
• The study design is appropriate for the regulatory question being addressed.

2. Use Cases

The guidance identifies several use cases where de-identified RWE is particularly appropriate:

• External Comparator Arms: Using RWE to construct a control group for single-arm clinical trials, particularly in rare diseases or oncology.
• Label Expansion: Supporting new indications or patient populations for already-approved products.
• Post-Market Safety Surveillance: Monitoring for safety signals in real-world populations.
• Medical Device Performance: Demonstrating real-world performance of devices post-approval.

3. Documentation Requirements

While the bar for data transfer has been lowered, the bar for documentation remains high. Sponsors must provide:

• A detailed description of the data source(s), including provenance, coverage, and known limitations.
• A study protocol or statistical analysis plan pre-specified before data analysis.
• Clear documentation of the de-identification process.
• Sensitivity and subgroup analyses to assess robustness.
• A commitment to provide IPD upon request if the FDA determines it is necessary for verification.

Why This Matters

This guidance has significant implications for sponsors, data providers, and the broader healthcare ecosystem.

Faster Time to Submission

Negotiating data sharing agreements and managing secure data transfers can add months to RWE study timelines. By accepting de-identified data, the FDA removes a significant logistical bottleneck.

Broader Data Access

Some data holders—particularly health systems and international registries—have been reluctant to share IPD due to privacy concerns and internal policies. The acceptance of de-identified data may open doors to new data sources that were previously off-limits.

Reduced Privacy Risk

Any transfer of patient-level data carries privacy risks, even with robust protections. De-identified, aggregate data significantly reduces this risk, making RWE more palatable to patients, providers, and data custodians.

Cost Savings

The infrastructure required to securely transfer and manage large IPD datasets is expensive. De-identified data submissions are typically smaller, simpler, and cheaper to prepare and review.

Use Cases: Putting the Guidance into Practice

Oncology: External Control Arms

One of the most immediate applications of this guidance is in oncology, where single-arm trials are common, especially in rare cancers. Sponsors can now more readily use de-identified data from cancer registries (e.g., SEER, NCDB) or EHR-derived datasets to construct external control arms, benchmarking their experimental therapy against real-world outcomes for standard of care.

Rare Diseases: Natural History Studies

For rare diseases, prospective controlled trials are often infeasible due to small patient populations. De-identified data from patient registries and advocacy group databases can provide crucial natural history data to contextualize trial results.

Medical Devices: Post-Market Surveillance

Device manufacturers can leverage de-identified data from hospital procedure databases and device registries to monitor real-world performance, detect safety signals, and support label expansions.

Practical Steps for Sponsors

Based on the new guidance, here are the key actions sponsors should take now:

1. Audit Your RWE Strategy

Review your current and planned RWE studies. Identify which ones could benefit from the de-identified data pathway. Consider whether data sources that were previously inaccessible might now be within reach.

2. Strengthen Data Provenance Documentation

The guidance places heavy emphasis on understanding and documenting data sources. Invest in building robust data provenance documentation, including data dictionaries, quality metrics, and limitation disclosures.

3. Pre-Specify Everything

The credibility of RWE analyses depends on pre-specification. Ensure that study protocols and statistical analysis plans are finalized and ideally registered publicly before data analysis begins.

4. Engage with the FDA Early

If you plan to use de-identified RWE for a regulatory submission, consider requesting a Pre-Submission meeting with the FDA to discuss your approach. Early alignment can prevent costly surprises later.

5. Build Flexibility for IPD Requests

Even though de-identified data is now acceptable, be prepared to provide IPD if requested. Ensure your data sharing agreements and internal processes allow for this contingency.

The Future of RWE in Regulatory Decision-Making

This guidance is part of a broader trend toward the integration of RWE into the regulatory fabric. We expect to see:

• Increased use of RWE for accelerated approvals: Particularly in areas like oncology, rare diseases, and infectious disease outbreaks.
• More RWE-based label expansions: Sponsors will increasingly use post-market RWE to support new indications without the need for new RCTs.
• Harmonization with international regulators: The EMA, PMDA, and other agencies are watching the FDA's approach closely. Global harmonization on RWE standards would benefit multinational drug development programs.
• Growth of the RWE ecosystem: Data providers, analytics companies, and RWE-focused CROs will continue to grow as demand for high-quality RWE increases.

Conclusion

The FDA's new guidance on de-identified Real-World Evidence is a significant step forward in the agency's journey toward evidence-based, patient-centric regulation. By removing barriers to the use of RWE, it opens new pathways for sponsors to bring innovative treatments to patients faster and more efficiently.

However, with greater flexibility comes greater responsibility. Sponsors must ensure that their RWE studies are designed rigorously, documented transparently, and analyzed robustly. The bar for evidence quality remains high—what has changed is the logistics of how that evidence is shared with regulators.

For organizations that invest in building strong RWE capabilities now, this guidance represents a significant competitive advantage. The future of drug development is real-world, and the future is now.

Stay informed on regulatory developments that matter. Subscribe to the CTDSU newsletter for weekly updates and expert analysis.