← All explorations Design Exploration · Heuristic Evaluation

MyChart Mobile —
What's working
and what isn't

Heuristic Evaluation · iOS App · August 2026

Eight findings across five flows — evaluated as a real patient, not a researcher with a protocol.

Why this app? I use MyChart as a patient managing an ongoing health condition. The evaluation came out of genuine frustration. I reviewed it against Nielsen's 10 Usability Heuristics and foundational UX principles, documenting eight distinct findings across five key flows.

4 High
2 Medium
2 Low
Finding 01
Medium

Patient selection ignores family hierarchy

Mental Model Alignment

The app opens with three equal-weight circles — me and my two children. But I'm the account holder. My children are dependents I manage on their behalf, not co-equal users. The flat structure requires me to consciously select myself every time, adding cognitive friction at the very first step.

Opportunity: A primary account view with dependents shown as a secondary tier would match how families actually think about shared healthcare management.
Patient selection screen showing three equal-weight profile circles

Three equal-weight profiles — the account holder has no visual priority over dependents

Finding 02
High

Home screen treats everything as equally urgent

Information Hierarchy · Cognitive Load

The home screen surfaces medical results, billing, admin prompts, and promotional content — all in the same card format with the same visual weight. A patient logging in to check a test result has to scan through everything to find what matters.

Opportunity: Content should be tiered by urgency: medical (test results, upcoming appointments) above administrative, with promotional content clearly separated or optional.
MyChart home screen showing mixed-priority cards with identical visual weight

Admin prompt and billing card share the same card format as test results — no visual triage

Finding 03
Medium

Shortcut icons look like navigation tabs — but aren't

Affordances & Signifiers · Consistency

Six shortcut icons appear at the top of the home screen. As you scroll, they collapse into smaller floating icons that persist as you move through content. The consistent icon-plus-label format within a sticky horizontal strip is the visual language of a tab bar — a pattern that conventionally signals in-place view switching, not navigation to a new screen. But tapping these opens entirely new screens, which breaks the expected interaction model.

Opportunity: If tabs are introduced elsewhere (e.g. for notification categories), the shortcuts need to be visually differentiated — different position, style, or interaction pattern.
Scrolled home screen showing floating shortcut icons that visually resemble a tab bar

Shortcuts collapse into floating icons as you scroll — tab bar pattern, but they navigate rather than switch views

Finding 04
Low

Lab order cards are verbose by default

Progressive Disclosure · Task-Oriented Design

Upcoming test reminders display the full instruction text, the panel name, clinical instructions, and a download button — all expanded by default. The primary action is to download the document. Everything else is context.

Opportunity: Lead with the test name and download action. Expand instructions on tap for users who need guidance.
Upcoming Tests card showing full clinical instructions expanded by default

Full clinical instructions shown by default — the download button is the only action most patients need

Finding 05
High

Scheduling collects information before checking availability

Error Prevention · Efficiency of Use · Progressive Disclosure

In the Women's Health scheduling flow, the app asks for pregnancy status before checking whether any appointments are actually available — so if the answer turns out to be "No available times," the question served no purpose. The pregnancy prompt may have clinical logic behind it (it could affect which providers are shown), but that makes it a candidate for post-booking intake rather than a pre-availability gate. More broadly, collecting specialty-specific information upfront, before confirming a slot exists, prioritises data collection over the patient's actual goal.

Opportunity: Checking availability first — then collecting intake details once a slot is confirmed — would reduce dead-end friction. Specialty-specific questions are already expected as pre-visit forms; moving them there keeps the booking flow focused on the one thing the patient came to do.
Women's Health flow asking pregnancy status before checking availability

Pregnancy status collected before any availability is checked

Finding 06
High

No provider-first entry point in scheduling

Mental Model Alignment · Efficiency of Use

The scheduling flow eventually shows which provider will see you — but only after you've navigated through a specialty menu to reach them. There's no path that starts with "Who do you want to see?" For a follow-up with an established provider, you have to know their specialty category and navigate there yourself. And if your provider's specialty isn't listed in the menu — endocrinology, for example, isn't — the flow has no way to reach them at all.

Opportunity: A provider-first entry point — "Book with Dr. Katta?" — would better match how returning patients think about scheduling. Established provider relationships are already in the system; surfacing them as the default starting point would save steps and feel significantly more personal.
Specialty selection screen with no provider-first option

Specialty menu — no path to start from your established provider

Finding 07
High

"No available times" is a dead end

Error Recovery

After a multi-step scheduling flow, the failure state reads: "No available times. Try a different search or call us." The recovery option is a phone call — which is exactly what the app was intended to replace. No saved preferences, no nearby provider suggestion, no indication of when slots might open.

Opportunity: Offer meaningful recovery: save the search criteria, suggest an alternative provider with availability, or allow the patient to set a notification when a slot opens.
Schedule appointment screen showing No available times with no useful recovery path

Dead end — after a multi-step flow, the only recovery option is a phone call

Finding 08
Low

Lab results are siloed by test order, not by blood draw

Mental Model Alignment · Information Architecture

MyChart handles grouped panels well — a lipid panel shows LDL, HDL, triglycerides, and cholesterol together in one trend view. But results from separately ordered tests live in isolation: HbA1c has its own trend, vitamin B12 has its own, vitamin D has its own. As someone tracking multiple markers across blood draws, I find myself navigating back to the test results list, locating each result individually, and opening it in a separate screen. There's no unified way to see all my blood work in one place.

Opportunity: A blood draw summary view — not a single merged chart, but a dashboard of all lab results navigable from one screen — would match how patients think about their lab work. Individual trend charts can stay as they are; the missing layer is a way to scan all results without navigating back and forth each time.
Result Trends screen showing a single glucose trend chart in isolation

One result at a time — no way to see all blood draw results without navigating back to the list

Where to focus

Four of the eight findings are in one flow.

The appointment scheduling flow compounds. A patient trying to book an appointment hits wrong-order questions, a missing provider relationship, and a dead-end failure state — all in a single task. Fixing the sequence and the recovery state would meaningfully reduce the number of patients who give up and call instead.

The home screen information hierarchy is the second priority: it affects every session, not just scheduling attempts. Medical content needs a clear visual priority over admin and promotional content.

→
Competitive audit
How do other patient-facing and scheduling-first apps handle the established provider relationship? The question isn't what's technically possible — it's what mental model patients actually bring to the scheduling experience.
→
Secondary research
Understand whether calling instead of using the app is a common pattern — and what triggers the switch. If enough patients default to the phone, that's a systems problem, not a UI problem.
→
Redesign concept
A provider-profile driven scheduling flow — browse available doctors in a specialty, read their bios and focus areas, choose who you want to see, then pick a time from their calendar. Specialty-specific clinical questions move to post-booking intake forms, where they already belong.
More explorations →