NHS physicians describe EHR interoperability as basic, fragmented, and clinically costly
A cross-sectional survey of NHS physicians paints a blunt picture of electronic record sharing across England: interoperability remains limited enough to interfere with everyday care delivery, not just administrative convenience. In this study, physicians described systems that often allow them to read information from within their own organisation but not meaningfully edit or exchange data across settings, leaving the record incomplete at the exact moment a decision has to be made. The practical effect was not abstract frustration. It reached the bedside, the clinic room, and the discharge pathway.
A total of 636 physicians took part, and 218 completed the survey in full. Responses were gathered online between June and October 2021 using convenience sampling and snowball recruitment through contacts at Health Education England deaneries and the Royal College of General Practitioners. The design was descriptive, but the clinical signal was unmistakable: interoperability was viewed as rudimentary across much of the NHS, with the sharpest criticism coming from secondary care specialties, physicians with less than 1 year of EHR experience, and those who rated their own EHR skills lower.
Limited data exchange slowed decisions and pushed patients through a slower pathway
Poor interoperability affected care in ways physicians could see immediately. Hospital stays were prolonged, consultation times became longer, and repeat investigations were frequently needed when prior results could not be accessed or trusted in full. In a busy clinical environment, each of those failures has a downstream cost. A missing scan report can delay discharge. A missing laboratory trend can trigger another blood draw. A missing clinic letter can force a fresh history from scratch, with all the uncertainty that introduces into diagnosis and treatment planning.
The pattern matters because interoperability is not just a technical feature hidden in the background. It determines whether a doctor can verify medication changes, review prior imaging, confirm specialist recommendations, or reconcile a patient’s story against objective data from another site. When records stay trapped in separate silos, clinicians spend more time recreating information that already exists, while patients wait longer for care that should have been faster and safer.
Limited interoperability also hampered communication between providers. In practice, that means clinicians cannot always assume that a decision made in one part of the system will be visible, legible, and actionable in another. A discharge plan may not reach community teams cleanly. A referral may arrive without the context that would have made it useful. A follow-up clinician may enter the consultation already working with partial information, which is a poor foundation for high-stakes care.
Safety risk emerged from missing context, duplicated tests, and fragmented oversight
Physicians perceived the interoperability gap as a threat to patient safety, and that concern is clinically credible. Safe care depends on continuity of information. When systems cannot exchange data effectively, the risk of medication discrepancies rises, as does the chance that abnormal results will be missed, duplicated, or acted on too late. Fragmented records also make it harder to see the trajectory of illness across services, which is especially important for patients with complex chronic disease, multiple admissions, or care that crosses primary and secondary settings.
The survey did not measure adverse events directly, and it did not quantify how often safety incidents occurred. Even so, the reported experience points toward a familiar mechanism of harm: incomplete information creates uncertainty, uncertainty drives defensive testing and conservative decisions, and both can slow care while increasing the opportunity for error. In that sense, the safety problem sits inside the workflow itself. The more fragmented the record, the more likely clinicians are to compensate with extra investigations, repeated questioning, and slower clinical decisions.
Physicians with shorter EHR experience and lower self-reported skill levels were more likely to hold negative views of interoperability. That finding suggests an important operational issue. When digital systems are difficult to navigate, poorly connected, or inconsistent across services, the burden falls hardest on clinicians who have had less time to develop workarounds. Senior familiarity may blunt some of the friction, but it cannot replace truly interoperable infrastructure.
Efficiency losses carried real consequences for staff time and patient flow
The efficiency findings were direct and pragmatic. Longer consultations are not a minor inconvenience in a stretched health service; they accumulate across clinics, reduce throughput, and make it harder to absorb urgent demand. Repeat investigations likewise consume staff time, laboratory capacity, and imaging resources that could have been used elsewhere. Even when those extra tests are ordered in the name of caution, they often reflect system failure rather than clinical need.
For physicians working in secondary care, where patients often arrive with more complex histories and care spans multiple teams, the limitations were especially visible. Interoperability failures are likely to be felt most sharply in these settings because the clinical stakes are higher and the number of handoffs is greater. A single missing document can delay a discharge decision, interrupt a treatment sequence, or force a patient to remain in hospital while information is chased manually.
One sentence captures the operational reality: incomplete records slow medicine down.
Policy solutions will have to move beyond local workarounds
The findings point toward a straightforward policy implication. If the NHS wants digital records to support safer and more efficient care, interoperability cannot remain partial, local, or dependent on individual clinician effort. Systems need to exchange usable information across organisational boundaries, and they need to do so in a way that is reliable for front-line practice, not just compliant on paper.
The study also suggests that implementation support matters. Physicians who reported lower EHR skill levels were more likely to view interoperability negatively, which implies that usability, training, and interface consistency may influence how well systems perform in real clinical settings. A technically connected network can still fail clinicians if information is buried, inaccessible at the right time, or too cumbersome to retrieve during a consultation.
The survey relied on convenience and snowball sampling, so it cannot claim to represent every NHS physician equally. Responses were also self-reported, which means the study captures perceived impact rather than independently measured delays, errors, or cost savings. Even with those limits, the message from clinicians was coherent: poor EHR interoperability is not a theoretical systems issue. It shapes waiting times, consultation burden, repeat testing, and the safety margin around patient care.
For a health service handling growing data volume and complexity, the next step is not simply more digital documentation. It is better connection between the systems already in use, so that clinical information can move with the patient instead of stalling at organisational borders.