The NIH's All of Us electronic health records program just got a new tool for closing a stubborn data gap: piggybacking on clinical data-sharing networks originally built to help doctors coordinate patient care, rather than to support research. The agency announced the shift as part of a data release on June 30, offering a potentially easier path to gather real-world medical records for researchers.
A Precision Medicine Program Built on Participant Data
Since 2018, All of Us has enrolled nearly 750,000 participants in what is one of the most ambitious precision medicine research efforts run by the National Institutes of Health. Participants contribute genome sequences, Fitbit activity data, family health history surveys, and electronic health records that can span an entire lifetime of care — all made available to researchers studying disease patterns, treatment outcomes, and genetic risk factors.
Why Electronic Health Record Gaps Are a Problem
The program's real-world medical records are meant to give researchers a fuller picture of participant health beyond what surveys and wearables can capture. But those records often arrive with major holes. Even though 98 percent of All of Us participants agree to share their electronic health records for research, more than 300,000 of them have no EHR data at all sitting in the program's database.
That gap matters because incomplete records limit what scientists can learn. A participant's genome and lifestyle data are far more useful when they can be linked to actual diagnoses, medications, and lab results over time — the kind of longitudinal detail that reveals how diseases develop and how treatments perform in the real world.
How the New Data-Sharing Approach Works
Rather than waiting for hospitals and clinics to individually transmit records, All of Us is now drawing on clinical data-sharing networks that already move patient information between providers to coordinate day-to-day care. Tapping into that existing infrastructure lets the program pull in records for participants who consented to share but whose data never showed up through traditional channels.
It is a notable shift in strategy: instead of building new pipelines from scratch, the program is repurposing systems designed for clinical care coordination and pointing them toward research. That kind of infrastructure reuse could make it meaningfully easier to close data gaps at scale, without requiring every hospital system to sign a new research-specific data-sharing agreement.
Why It Matters for Medical Research
All of Us was designed from the start to reflect the diversity of the US population, with an emphasis on enrolling participants from communities that have historically been underrepresented in biomedical research. Missing health records disproportionately undermine that goal, because gaps in data can quietly erase the very populations the program is trying to include.
A more complete real-world dataset also strengthens the program's core promise: giving researchers, drug developers, and public health officials a resource to study how diseases and treatments play out across genetically and demographically diverse groups, rather than the narrower populations that have historically dominated clinical research. For readers following how large datasets are reshaping medicine, our health coverage and science coverage track similar developments as they unfold.
All of Us EHR Data FAQ
What is the All of Us Research Program?
All of Us is a National Institutes of Health initiative that has enrolled nearly 750,000 participants since 2018 to build one of the largest and most diverse health datasets in the world, combining genetic, survey, wearable, and medical record data.
Why don't all participants have electronic health records in the database?
Even though 98 percent of participants consent to share their records, technical and logistical barriers between health systems have historically left more than 300,000 participants without any EHR data on file.
How does the new approach fix that gap?
The program is now drawing on clinical data-sharing networks that hospitals already use to coordinate patient care, using that existing infrastructure to pull in missing records rather than building separate research-only pipelines.
