For years, one of the practical limits on using real-world evidence in regulatory submissions was a straightforward data access problem: the FDA expected sponsors to submit individually identifiable patient-level data when incorporating real-world data (RWD) into applications, which made it difficult to draw on the large, de-identified datasets that have become increasingly central to clinical research.
In December 2025, the FDA updated its guidance on the use of real-world evidence (RWE) to support regulatory decision-making for medical devices, allowing sponsors to submit RWE without patient identification data. The change means that data derived from sources such as electronic health records (EHRs), insurance claims, and disease registries can now be used to support certain medical device submissions without requiring the underlying identifiable patient information.
Previously, only 35 drugs, biologics, or vaccines had incorporated RWE into their applications since 2016, compared with more than 250 premarket device authorizations that included RWE during the same period — though even the rate of RWE-supported device authorizations had plateaued in recent years. The updated guidance is designed to address the practical constraints that kept that number low.
What changes — and for whom
The immediate impact of the December 2025 guidance is specific to medical devices. For drugs and biologics, the FDA has indicated it intends to consider updating parallel guidance in the future, but until then, drug sponsors should continue engaging the FDA early and on a case-by-case basis when proposing RWE approaches that may not include full participant-level data.
The distinction matters for drug developers paying attention to where this is heading. The guidance does not relax data quality standards — it changes what form that data needs to take at the point of submission. Sponsors are still expected to demonstrate data provenance, document preprocessing steps, and address potential sources of bias with the same rigor that has historically been applied to traditional clinical trial data.
"The FDA's update enhances clarity on the use of real-world evidence constructed from diverse de-identified patient data,"Jen Lamppa told DDN. Lamppa is the Vice President of Commercial Strategy at Inovalon, and has spent more than two decades applying real-world data across more than 100 clinical studies. "This guidance also reinforces the bar for data reliability and robustness, regardless of whether identified or de-identified, with demonstrable data provenance, quality checks, processes, and pipelines."
Implications for trial design
One of the more immediate practical implications involves external control arms and hybrid trial designs. Access to large de-identified datasets makes it more feasible to construct external control arms that reflect standard-of-care patient populations without requiring those patients to be prospectively recruited and consented in the traditional sense. That has the potential to reduce trial size, shorten timelines, and lower development costs — particularly in disease areas where patient recruitment is inherently difficult.
Lamppa pointed to post-market commitments and indication expansion as areas likely to see the most near-term impact, where evidence on broader and more heterogeneous patient populations is often needed and large de-identified datasets are well suited to fill that gap. The tradeoffs, however, are real. Real-world datasets carry their own methodological challenges — incomplete fields, inconsistent coding, variable follow-up, and selection bias among them — that sponsors will need to address proactively.
"Real world data are in many ways more challenging to handle than traditional study data," Lamppa said. "Regulators expect sponsors to proactively address these complexities with established data standards that prioritize traceability and confront bias from study onset."
A signal, not a replacement
It would be easy to read the FDA's updated guidance as a step toward phasing out traditional randomized controlled trials, but that framing overstates what the policy change actually does. Randomized controlled trials remain the evidentiary standard for pivotal submissions, and the guidance is better understood as formalizing a complementary role for RWE rather than replacing it.
What the update does signal is that the FDA is increasingly willing to engage with the practical realities of how clinical evidence is generated at scale — and that drug and biologic sponsors would do well to start building the infrastructure and methodological rigor now to be ready if and when similar guidance arrives for their modalities.
"De-identified RWD can be seen as a practical way to accelerate evidence generation and close gaps that traditional data approaches can't efficiently address," Lamppa said. "Broader reliance on de-identified [data] will depend on its ability to demonstrate consistent quality, reproducibility, and applicability at scale."
