Patient portals have some of the worst UX of any product category. They are built to comply with regulations, serve the administrative needs of the organisation, and check the 'digital health' box — not to help patients manage their health. When a regional healthcare network came to us with a portal that 23% of registered patients actually used, we were not surprised. We were ready.
Diagnosing the Problem
The portal in question had been built five years earlier by an EHR vendor and had received no significant UX investment since. It had 14 top-level navigation items. Appointment booking required navigating through four nested menus. Lab results were displayed as raw data tables — columns of numbers with no reference ranges, no plain-English explanation, and no guidance on what to do with the information.
We ran three weeks of user research before touching the design: moderated usability sessions with 22 patients across age groups, a survey of 340 registered portal users, and an analysis of the support call transcripts from the preceding 6 months. The data was unambiguous: patients were not using the portal because they could not accomplish what they came to do.
Research Finding
67% of portal support calls were from patients trying to complete one of three tasks: booking an appointment, viewing a test result, or messaging their care team. All three tasks required more than 4 navigation steps. One required 9.
The Information Architecture Redesign
The 14-item navigation was replaced with a dashboard model centred on three primary tasks: Book an Appointment, My Results, and Message My Team. These three tasks accounted for 78% of all portal sessions in the existing system. Making them the starting point, rather than buried destinations, was the single highest-impact structural change.
The dashboard also introduced a 'What needs your attention' panel — surfacing upcoming appointments, unread test results, and pending prescription refills without the patient needing to navigate to find them. For elderly patients and patients managing multiple conditions, this single panel replaced the need to visit multiple sections of the portal on every login.
Rewriting Lab Results
The lab results redesign was the most clinically sensitive part of the project. We worked closely with the network's clinical team to develop a plain-English explanation framework for the 40 most common test types — one that would give patients useful context without overstepping the role of clinical interpretation.
Each result now displays: the test name, the patient's result, the reference range, a plain-English status indicator (Normal, Monitor, Discuss with your doctor), and a brief explanation of what the test measures. The 'Discuss with your doctor' status links directly to the secure messaging feature. Support calls from patients who did not understand their results dropped by 71% in the first month post-launch.
The Appointment Booking Flow
Appointment booking was redesigned from a 9-step process requiring selection of department, clinic, service type, provider, appointment type, date, time, reason, and confirmation — to a 3-step flow: who do you want to see, when are you available, confirm. The system handled the mapping to department and service type invisibly, based on the patient's existing care relationships.
For patients without an established provider relationship, a brief triage question ('What do you need help with today?') routed them to the appropriate service. This prevented the most common booking error in the original system — patients selecting the wrong service type and arriving at appointments that couldn't serve their actual need.
Portal active usage rate
Before
23%
After
61%
Appointment booking completion
Before
41%
After
91%
Support calls: portal questions
Before
340/month
After
163/month
Patient satisfaction score (portal)
Before
3.1/5
After
4.6/5
The Lesson From This Engagement
Healthcare UX is often treated as a special category where the normal rules of good design do not apply because of regulatory complexity and legacy system constraints. This engagement demonstrates the opposite: when you apply the same standards of clarity, task focus, and user empathy that you would to any product, even the most entrenched legacy experience can be fundamentally improved.
We did not replace the underlying EHR system. We did not change the data model. We changed how information was structured, labelled, and presented — and the patients who used the portal stopped needing to call for help understanding it.
“We expected the engagement to be primarily about compliance and accessibility. It turned out to be primarily about respect — designing an experience that respected the patient's time, intelligence, and anxiety about their health. That framing changed everything.”
— Director of Digital Health, Regional Healthcare Network