Clinical vs Patient Interfaces

A blood-test result is a single entry in a laboratory system, but it reaches two people in very different situations. A clinician may meet it between many other tasks, one row among dozens. A patient may open it alone at home, sometimes before anyone has had a chance to explain it. In the United States and in Australia, recent changes have moved towards patients seeing results at about the same time as, or before, the clinician. The screen then becomes the only interpreter present.

Two audiences, two jobs

A clinical interface supports work: scanning, deciding and acting across many patients, alongside ordering and documentation. A patient interface supports understanding: what this means for me, whether it is serious, what happens next. The data can be identical; the questions asked of it are not. The NHS service manual's design principles point in this direction by asking services to consider people's emotional, physical and technical needs and the wider context of use.

An electronic health record (EHR) is a provider's digital patient file; a patient portal is the patient-facing website or app showing part of it.

What clinician screens have to cope with

Clinician interfaces are dense on purpose and often interrupted. Alert fatigue is the pattern in which people stop responding to alerts because too many arrive, including unhelpful ones. A 2022 stewardship review in Applied Clinical Informatics reports low acceptance of interruptive pop-up alerts, in the range of 4 to 11 percent, and argues that poor alerts weaken trust in all of them. A 2026 systematic review of 22 studies found that alert fatigue is measured inconsistently, and only 3 of the 22 reported whether overrides were appropriate.

Labelling and navigation matter as much as alerts. In a study of a tablet interface for remotely monitoring patients' vital signs in a hospital care unit, 10 staff members found 37 usability problems in the original screens, mostly unclear labels and confusing navigation. After a redesign, mean scores on the System Usability Scale, a standard 0 to 100 questionnaire, rose from 68.5 to 89. The sample is small, but it shows that clarity is a safety matter for clinicians.

What patient screens have to do

Patient portals have often shown lab results in tables built for clinicians. A 2021 JMIR Human Factors study surveyed 203 people, interviewed 13 patients and tested a prototype that paired graphics with a plain-language takeaway message. The pilot had only 8 participants, all of whom judged it an improvement on existing portals, so it shows a direction rather than proof. Some participants also worried that even mild phrasing such as "not optimal" could cause needless anxiety, which shows that word choice is part of the interface.

How well do patients cope in practice? A 2026 analysis of a 2024 US national survey found that 68.6 percent of adults who had been offered online record access and had a test in the past year viewed immediately released results before hearing from their clinician. Of those who viewed results, 6.6 percent rated their own understanding as poor. That is self-reported understanding, not a test of accuracy.

Timing changes the design problem

The US federal rule implementing the 21st Century Cures Act requires that patients can electronically access their electronic health information at no cost, and it restricts practices that block that access, with ten exceptions. At one academic medical centre, the share of previously delayed results that patients viewed before their clinicians rose from 10.4 percent to 40.3 percent after compliance, and the median number of patient messages sent soon after viewing rose from 77.5 to 146 per day. The authors cautioned that benefits could be offset by effects on patient well-being and clinician workload; this is one centre.

Australia has moved in a similar direction. According to Australian Digital Health Agency fact sheets for this change, most pathology reports in My Health Record became viewable by patients as soon as they were uploaded from October 2025, replacing a 7-day delay. Some categories, such as anatomical pathology, cytopathology and genetic tests, keep a 5-day delay, and healthcare providers continue to have access to all results once uploaded. A Medical Journal of Australia commentary sets out both sides: the delay gives the GP time to interpret results, while patients have a right to their own information and urgent findings can be missed during a wait. Either way, a patient interface cannot assume a conversation has already happened.

Two examples

First, a result outside its reference range, meaning the span a laboratory treats as typical. The clinician's view is a compact, sortable list across patients, with flags and trends, ordered by what needs attention. The patient's view uses the plain name of the test, shows where the value sits against the range, explains in careful wording what "outside the range" does and does not mean, and says who to contact. A Showcase idea that sets a clinician dashboard beside a patient trend chart explores this contrast.

Second, remote patient monitoring, where a device such as a home blood-pressure cuff sends readings to a care team. The patient needs confirmation that the reading arrived and a sense of its meaning. The care team needs a panel of patients ordered by need. In a JMIR Human Factors study of a programme for cancer patients after hospital discharge, 30 clinicians responded (31 percent of those invited); 27 percent agreed the dashboard was well organised, 81 percent said the programme did not reduce their workload, and repeated alerts for patients reporting stable symptoms led to unnecessary verification calls. The sample is small and self-selected.

Related Showcase card

• Humanising the number - the closest built example so far of the same underlying value read two ways, patient-facing without a clinician alongside it. The clinician-dashboard and patient-trend-chart pair this article was written to sit next to is not built yet.