A full UX prototype for a patient-facing appointment and triage app, built from discovery through to a tested, clickable Figma prototype in 6 weeks.
CareTrack came to us with a clear problem: NHS triage pathways are confusing, stressful, and poorly designed for the patients who need them most. Their founders — a GP and a health-tech entrepreneur — had watched patients arrive at A&E with conditions that could have been managed remotely, simply because the existing tools made it too hard to understand what care they actually needed.
The ask was to design a patient-facing mobile app that would guide users through a symptom assessment, help them book appropriate appointments, and surface the right NHS resource (GP, urgent care, 111, A&E) based on their responses — without ever making them feel like they were talking to a call centre script. Six weeks. Full prototype. Investor-ready.
Maria led a three-part discovery phase: stakeholder interviews with the two founders to extract the clinical logic and commercial constraints; a desk research sweep through NHS Digital guidelines, NICE pathways, and existing triage tools (111 Online, Babylon, Livi); and a round of five unmoderated user interviews with people aged 25–65 who had used NHS services in the past 12 months.
The research surfaced three consistent pain points. First, patients didn't trust their own judgement — they needed the app to validate their concern, not just redirect them. Second, the language used in existing tools was clinical and distancing; people switched off when they saw terms like "presenting complaint." Third, the booking flow in most NHS tools required too much prior knowledge (knowing your GP surgery's system, having your NHS number to hand) and punished people who didn't have it.
"I always feel like I'm being stupid for calling, or that I should have just gone in. I never know if I'm doing the right thing." Research participant, 34
These three insights shaped every design decision that followed. The triage flow had to feel conversational, not clinical. The language had to be warm and direct. And the booking pathway had to work with as little information as possible, with optional extras rather than mandatory fields.
Mapping the triage logic was the most technically complex part of the project. NHS pathways aren't linear — they branch based on age, existing conditions, symptom severity, and time of day (some routes are only available during GP hours). We worked with the CareTrack GP founder to document every decision node and build a logic map that the design could sit on top of.
The resulting IA had four top-level entry points: Start a symptom check, Book an appointment, View my care history, and Urgent help. The symptom check followed a branching card-based format — one question per screen, clear progress indicator, always an option to say "I'm not sure." At the end of each pathway, the app surfaced a recommended action with a plain-English explanation of why, and a one-tap route to the next step.
We ran the logic map through a paper prototype session with three participants before moving to Figma — catching two critical branch errors (a respiratory pathway that skipped the asthma flag, and an abdominal pain route that didn't account for pregnancy) before they made it into the design.
The visual language was built to feel trustworthy without feeling like a hospital. We kept the palette to a calm blue-grey and white system, with a single warm accent used only for CTAs and reassurance states. Typography was Inter throughout — legible at small sizes, friendly without being frivolous. Every element followed WCAG AA contrast requirements as a baseline.
The 48-screen prototype covered the full user journey: onboarding and account creation (with optional NHS login), the complete symptom triage flow across four symptom categories (respiratory, musculoskeletal, mental health, general illness), appointment booking with GP and urgent care, and a care history view. Each screen was built as a Figma component with clearly defined states — default, active, error, empty.
We also designed a set of micro-interactions that made the triage feel less like a form and more like a conversation: a subtle breathing animation on the loading state, progress steps that confirmed backward as well as forward movement, and a soft checkmark completion state that gave users explicit validation before redirecting them.
We ran two rounds of moderated testing — five participants in round one, four in round two after iteration. Participants were recruited to match the target demographic: adults over 25 who were not healthcare professionals and who had used NHS services at least twice in the past year.
Round one identified the biggest friction points: the account creation screen had too many required fields before any value was delivered; the "severity slider" in the symptom check was being misread as a rating rather than a current-state selector; and the recommended outcome screen wasn't giving enough reassurance before redirecting to 111.
We made targeted changes between rounds: moved account creation to post-triage (guests first), replaced the slider with a three-option card selector (Mild / Moderate / Severe), and rewrote the outcome copy to lead with validation before the action. Round two showed a 60% reduction in observed task friction — participants moved through the full triage flow without pausing or second-guessing their choices.
The prototype was delivered on week six — 48 screens, a full component library, a logic map document, and a testing report. CareTrack used it to close their pre-seed round and are currently in development with an NHS trust pilot planned for Q3.
The work demonstrated something we believe about health-tech design: the clinical rigour and the human warmth aren't in tension. You don't have to choose between accurate and kind. With the right research, the right architecture, and careful writing, you can build something that genuinely helps people make better decisions about their own health — without making them feel judged for not knowing the answer.
"The prototype was investor-ready and patient-ready. Maria's research grounding gave the whole project a credibility that other agencies simply couldn't have matched." Rachel T., CareTrack Health
We bring research depth and design craft to complex problems. Let's talk about what you're building.
Start the conversation → See more work →