FDA signals broader acceptance of de-identified real-world data in regulatory reviews

Updated FDA guidance will allow certain medical device applications to rely on real-world evidence (RWE) derived from de-identified data sources, potentially expanding the role of large real-world datasets in regulatory decision-making.

The US Food and Drug Administration (FDA) has announced that certain medical device applications may now rely on RWE derived from de-identified data sources, removing a long-standing barrier to broader regulatory use of real-world data (RWD).

Under updated guidance for specific types of medical device submissions, the FDA will no longer require that identifiable individual patient data always accompany RWE included in marketing applications. The agency has also signaled that it intends to consider similar updates to its guidance for drugs and biologics.

Since 2016, RWE has been included in regulatory applications for 35 drugs, biologics, or vaccines. Uptake has been more substantial in the medical device space, with more than 250 premarket authorizations incorporating RWE over the same period. However, the FDA has noted that growth in RWE-supported device approvals has slowed in recent years. One contributing factor has been the agency’s historical expectation that RWE submissions include confidential, patient-level data, limiting the usability of many large-scale RWD sources, even where they contain robust information on outcomes, treatment patterns, and safety.

In response to growing calls from manufacturers and data scientists to modernize these requirements, the FDA will now consider RWE derived from de-identified RWD sources, including patient registries, insurance claims data, and electronic health records. The strength and relevance of submitted evidence will be assessed on an application-by-application basis. The updated approach reflects the agency’s view that reliable and meaningful evidence can, in some cases, be generated from aggregated or de-identified data without access to identifiable patient records.

FDA Commissioner Marty Makary said the change removes unnecessary constraints on the agency’s ability to use existing data resources. “This common-sense reform will unlock access to vast databases like cancer and cystic fibrosis registries that contain critical insights about how treatments work in the real world,” he remarked.

By lowering the data access threshold for RWE submissions, the FDA has expanded the range of RWD sources that may inform regulatory decision-making. For manufacturers, this may reduce practical and legal barriers associated with data sharing, while improving the feasibility of using existing datasets. For regulators, the change could support broader evaluation of treatment performance across diverse populations and real-world settings, while retaining application-specific scrutiny of evidence quality and relevance.

Makary has also framed the shift as part of a broader reassessment of evidentiary expectations at the agency. In a recent interview, he noted that while FDA reviewers will continue to challenge and reject RWD that does not meet regulatory standards, prior expectations around data provenance and consent were often unrealistic for large-scale real-world datasets. He also pointed to the agency’s intention to move away from a default requirement for two pivotal clinical trials in every drug application, favoring instead one well-controlled, statistically robust trial, supported where appropriate by additional evidence. This he remarked “is going to allow for more innovation and less delays and lower R&D costs, which could translate to lower drug prices.”

The FDA’s acceptance of RWE derived from de-identified data may reflect a wider shift toward more flexible evidence generation. In 2023, the agency issued draft guidance clarifying when a single pivotal trial, supported by confirmatory evidence, may be sufficient. While the FDA has not explicitly linked the two policies, their alignment suggests increasing openness to alternative sources of credible evidence.

With the FDA expected to publish final guidance on the “Use of Real-World Evidence to Support Regulatory Decision-Making for Medical Devices”, the announcement raises the question of whether the long-awaited final version – released in draft form in December 2023 – may arrive sooner than expected, offering a timely development for those working in the medical device sector.