Providers See a Path for Patient-Generated Data to Enter the EHR
Physicians and information technology operations staff at Columbia University Irving Medical Center described a cautious but workable route for writing patient-generated edits back into the electronic health record. Their central concern was not whether patients should be allowed to contribute corrections, but how to do it without compromising data quality, provenance, or clinical trust.
The issue sits at the intersection of workflow, safety, and communication. In an environment where the EHR has become the operational record for diagnosis, treatment, and care coordination, even small errors can spread quickly through downstream decision-making. The interviews suggest that patient input may help close gaps in incomplete records, but only if the system preserves clinician oversight and clear data governance.
RealRisks Served as the Test Case for a Write-Back Workflow
The study used RealRisks, a Fast Healthcare Interoperability Resources-compliant web-based patient decision aid, as the model platform. Within that framework, patients can access selected EHR data and review, correct, or contribute patient-derived information when certain elements are missing. The design made the question concrete: if a patient sees an error or gap, should the corrected version flow back into the EHR rather than remain outside it?
Investigators conducted semi-structured interviews with 9 participants, including 4 physicians and 5 IT operations staff. The small sample size limits generalization, but the group still offered a useful view into the operational reality of adopting write-back functionality. Their responses were not uniform, which is unsurprising. Clinical documentation systems are built around authority, and patient-entered data challenges that longstanding hierarchy in a very practical way.
Accuracy, Not Autonomy Alone, Drove the Conversation
Participants generally recognized the need to improve EHR accuracy, but they framed that goal in relational terms rather than as a purely technical fix. Better records require provider-patient communication, not just a new interface or an additional button in a portal. In other words, patient participation may work only if clinicians remain involved in confirming what belongs in the chart.
A short sentence captured the operational tension: participants supported write-back, but not blind acceptance of every edit.
That caution mattered because patient-generated data can be useful for filling omissions, correcting outdated entries, or adding context that never made it into the encounter note. Yet the same channel could also introduce confusion if edits arrive without verification, if the system cannot distinguish a patient suggestion from a confirmed record, or if the workflow creates duplicate fields that clinicians then must reconcile manually.
Screening Requests Was Viewed as Essential to Data Quality
One of the clearest messages from the interviews was the need for provider screening of edit requests. Participants wanted existing mechanisms used for external data incorporation to serve as the template rather than inventing an entirely separate pathway. That preference points toward a conservative implementation strategy: leverage processes already designed to manage outside information, then adapt them for patient submissions.
This approach reflects a clinical logic familiar to anyone who has worked with imported lab results, outside records, or scanned documents. External data can enrich the chart, but only after someone confirms relevance and accuracy. The same principle would apply to patient-generated edits. Without review, the EHR risks becoming a repository of unvetted fragments rather than a reliable longitudinal record.
The participants also suggested a more structured submission process, including the option to upload test results when submitting changes. That detail is important because it shows how users may want to supply evidence along with the correction itself. A patient who identifies an error in a medication list or missing history item may also have supporting documentation ready to attach, which could reduce ambiguity and speed clinician review.
Interoperability Helps, but Workflow Still Decides Whether the System Works
Because RealRisks is FHIR-compliant, the study sits inside the broader push toward interoperability and patient access. Technical standards make write-back possible, but they do not resolve questions about accountability, triage, or final sign-off. The interviews make clear that implementation hinges less on whether data can be transmitted and more on who reviews it, how it is flagged, and where it lands in the record.
That is where IT operations staff become central. Their role is not limited to maintaining software; they shape the rules that determine whether a patient correction becomes a chart amendment, a pending request, or a separate informational note. The physicians in the study, meanwhile, appear to have viewed that oversight as necessary to preserve the reliability of the EHR as a clinical tool.
Equity Will Depend on Whose Perspective Is Still Missing
The investigators concluded that future work should include nursing staff, advanced practice providers, and patient perspectives. That next step is not optional. A write-back process that seems practical to physicians and IT staff may look very different to the clinicians who handle day-to-day triage or to patients with different levels of digital access, health literacy, or comfort challenging the record.
Equity concerns are baked into the problem. Patients who are more confident navigating portals may benefit first, while others could be left behind if the workflow assumes digital fluency or repeated follow-up. If the system is designed poorly, patient-generated corrections could become another source of friction for those with the least time, the least access, or the least ability to advocate for themselves.
The study does not resolve those questions, but it does define the boundary of what is currently plausible. Patient-generated data can enter the EHR in a useful way, yet only if providers retain a screening role, IT teams build controlled pathways, and future testing expands beyond physicians and operations staff. The next phase should determine whether that balance holds when nursing, advanced practice, and patient voices are placed inside the same workflow.