By the Phenomenon Studio product team
Why patient portals still sit unused even when HIPAA compliance is technically solved, and what separates a portal patients actually log into from one that only satisfies an audit.
As of 2026, most healthcare organizations in the US have already checked the box on patient portals. A federal data brief published this year found that 77% of patients were offered online access to their health records, up from 73% two years earlier. The harder number sits right next to it: only 65% of those patients logged in at least once. A portal that exists and a portal that gets used are two different products, and the gap between them usually comes down to design, not compliance.
Most teams building this kind of product start from the regulation and work outward: encrypt the data, log the access, lock down the endpoints, ship it. That gets a portal through a security review. It doesn’t explain why a patient who was handed working credentials still calls the front desk to ask about a lab result instead of opening the app. Mobile app design services built around that second question, not just the first one, tend to produce a noticeably different product.
What a patient portal has to do
A patient portal is not a smaller, patient-facing copy of an electronic health record. An EHR is built for someone trained to read it. A portal has to translate the same underlying data into language a person reads once, usually while worried about what it means. That includes lab values, medication lists, and visit summaries. Framing the portal as a stripped-down clinical record instead of its own product is the first mistake, and it’s usually invisible until real patients start logging in and calling with questions the interface should have already answered.
The information architecture problem is harder than it looks from a wireframe. A single portal often has to serve a patient managing one condition, a parent managing a child’s records, and an adult child managing a parent’s care, sometimes from the same login. Website design services scoped around a single generic patient persona tend to miss this branching almost entirely, because the demo account never has to juggle more than one identity.
Identity, proxy access, and the friction nobody budgets for
Authentication is where a lot of otherwise well-built portals lose patients before they see a single record. Identity verification exists for a real reason: a portal is handling protected health information, and getting it wrong is a breach, not a bug. But every additional verification step is also a point where a patient gives up and calls instead. The real design question is how to verify identity without turning every login into a small ordeal.
Proxy access makes this considerably harder, and it’s growing fast enough that it can’t be treated as an edge case anymore. Caregiver access to a patient’s portal more than doubled between 2020 and 2024, rising from 24% to 51% of patients with an authorized proxy on the account.
According to a 2024 ONC Health IT data brief, caregiver or proxy access to patient portals rose from 24% in 2020 to 51% in 2024, more than doubling in four years. (ONC Health IT Data Brief, “Individuals’ Access and Use of Patient Portals and Smartphone Health Apps, 2024”)
A portal built around a single-user assumption treats proxy access as a permissions afterthought, and it shows. Consent screens that don’t clearly state what a caregiver can see, revoke, or act on leave both parties guessing, which is exactly the kind of ambiguity that turns into a support call or, worse, a permission granted longer than the patient intended. A user experience design company that has shipped multi-party healthcare access before treats this as a core flow to design, not a checkbox to add once the single-user version ships.
Secure messaging and billing are where trust is won or lost
Two features inside most patient portals get far less design attention than the records view, and they’re often where a patient forms a lasting opinion of the whole product. Secure messaging is the first. A message thread that doesn’t clearly show whether a provider has seen the message, or how long a reply usually takes, leaves a patient anxious in a way a well-designed thread simply prevents. Web app development for a messaging feature that treats response-time expectations as a support-team problem, not an interface problem, tends to generate the exact complaints a good design would have headed off.
Billing is the second, and it’s frequently the ugliest part of an otherwise clean portal. Medical billing is already confusing on paper: multiple line items, insurance adjustments, dates that don’t obviously map to a visit the patient remembers. A portal that reproduces that confusion on a screen, rather than translating it, pushes patients straight back to a phone call with billing, which is the outcome the portal was supposed to prevent. Mobile app development company teams that treat billing as a data-display problem instead of a comprehension problem consistently underinvest here relative to how much patient frustration it generates.
Comparing the common ways portals get built
Most healthcare organizations choose between three broad approaches when a portal needs work, and each carries a different risk profile.
| Approach | Where it’s strong | Where it falls short |
| Default EHR vendor portal | Fast to enable, tightly tied to clinical data | Interface rarely built for a patient’s reading level |
| Fully custom build | Interface designed around real patient behavior | Slower to ship, needs ongoing compliance ownership |
| Custom layer over the EHR | Patient-facing experience redesigned without replacing clinical data source | Requires close coordination between two systems and two teams |
None of these is universally correct. A large hospital system with a locked-in EHR contract usually can’t justify a fully custom rebuild, but it can commission a custom layer that fixes the patient-facing experience without touching the clinical record underneath. A smaller practice or a digital-health startup building a portal from scratch has more room for a fully custom approach, provided the team understands that compliance ownership doesn’t end at launch.
What HIPAA-compliant design looks like beyond the backend
Encrypting data at rest and in transit is table stakes. So is logging access. None of it is visible to the patient using the product. HIPAA-compliant design also has a layer patients see directly. Consent screens need to state plainly what’s being shared and with whom. Session timeouts shouldn’t silently log a patient out mid-task. And account recovery flows need to verify identity without punishing a patient who simply forgot a password during a stressful week. UI UX design services that treat compliance purely as a legal review miss this visible layer almost every time, because a legal review checks whether the right disclosures exist, not whether a patient understands them on first read.
Website development company teams sourcing this work should ask directly how a prospective vendor handles the account recovery flow specifically, since it’s one of the more common places a compliant-on-paper design creates real friction. A recovery process that takes fifteen minutes and three phone calls technically protects the data. It also guarantees the patient won’t try the portal again for months.
Expert insight
Oleksandr Kostiuchenko, Marketing Manager at Phenomenon Studio, has found that a strong login-to-usage ratio correlates less with feature count and more with whether someone tested the account recovery and consent flows against patients who weren’t already comfortable with technology, not just internal staff who already understood the product. Skipping that step tends to hide the exact friction points that later show up as abandoned logins and repeat phone calls to the front desk.
Where the mobile experience earns or loses trust separately
A meaningful share of patients now reach a portal through a phone first, not a browser, which changes what “good enough” means for the interface. Mobile app development services scoped as a smaller, feature-reduced copy of the web portal tend to frustrate exactly the patients most likely to rely on the app as their primary way in, particularly for scheduling and secure messaging, the two features used most often between visits.
A mobile app development agency building this work needs the same compliance discipline as the web team, applied to a smaller screen with different interaction patterns. That means biometric login instead of a typed password, and push notifications that summarize a message without exposing protected health information in a lock-screen preview. It also means forms that don’t assume a full keyboard and a large display. Web development services quoted for the web portal and mobile design work quoted separately for the app should still map to one shared design system, so a lab result reads the same way regardless of which surface a patient opens first.
What to ask before choosing a partner for this work
A team evaluating vendors for a patient portal build should ask more specific questions than a general capability pitch answers. Ask a prospective web development agency to walk through a proxy access flow they’ve shipped, not just described in a proposal, since this is one of the areas that separates real healthcare experience from general product work. Ask how a website development agency handles account recovery specifically, and request a walkthrough rather than a written description.
Web design services quoted as a flat number for the entire build should raise a question, since scope in this category tends to shift once real compliance and usability findings surface mid-project. A website development company working in phases, with room to adjust after early testing, is usually a better sign than one promising a fixed price against a spec that hasn’t been tested with real patients yet.
UX design agency partners proposing to handle only the interface layer should explain clearly how they’ll coordinate with whoever owns the underlying data pipeline, since a portal redesign that changes the interface without understanding the EHR integration underneath it tends to produce inconsistencies within the first release. A web design agency should also be able to describe how it tests comprehension. Task completion alone proves little: a patient who technically finds the right button but misreads what a message says hasn’t been served well.
If branding work is part of the scope, ask whether the team works directly with branding companies or manages visual identity in-house, since a patient portal’s tone has to read as trustworthy and calm without feeling clinical and cold. A UX design agency that has designed for a nervous or unwell user before will usually have a clear answer to that question rather than a generic one.
How provider labels map to portal work
Search for help building this kind of product and the results blur together fast. Mobile app design services, website design services, and general web design services all claim overlapping territory, and the label on a proposal rarely predicts whether a team has handled protected health information before. Website design services quoted as a flat-rate package rarely account for the ongoing compliance review a portal needs after every content change. A generalist offering mobile app design services for consumer apps brings real interface skill that doesn’t automatically transfer to a proxy-access flow or a consent screen that has to hold up under a compliance review.
Website development company candidates split along a similar line. Some are strong at fast, templated builds for a practice’s public-facing marketing site. Fewer have shipped the harder, ongoing work a patient portal needs: account recovery, secure messaging, and a billing view that translates rather than just displays. General product experience isn’t enough on its own. A team should be able to point to at least one prior healthcare build before it gets seriously considered for portal work specifically.
Mobile app development company partners face the identical gap. A team confident with mobile app design services for a retail or fitness app is not automatically ready for biometric login, push notifications that can’t leak protected health information, and a scheduling flow that has to work for a patient managing a parent’s care from their own phone. A user experience design company evaluated for this work should be asked directly how many prior healthcare products it has shipped, not just how strong its general portfolio looks.
UI UX design services scoped narrowly to visual polish miss most of what determines whether a patient portal gets used. The interface work that matters here sits closer to a user experience design company’s core discipline than to a typical web design agency’s usual scope. That’s consent language, error recovery, and comprehension testing with real patients, not just visual polish. A web development agency asked to lead this kind of build should explain how it plans to involve that deeper research discipline, not just its own engineering timeline.
A useful gut check before signing anyone: ask how a candidate mobile app design services team would redesign the medication list specifically, and how a user experience design company on the same shortlist would test whether patients actually understand it. If a web development agency can’t answer the first question concretely and a UX design agency can’t answer the second, neither has done this specific kind of work before, regardless of how the rest of the pitch sounds. Teams that combine mobile app design services with genuine user experience design company depth under one roof tend to avoid the coordination gap that shows up when two specialists never compare notes on the same patient flow.
The measure that matters
A patient portal succeeds when the gap between “offered” and “actually used” closes, not when a security audit passes. Both matter, but only one of them is visible to the patient deciding whether to open the app again after a confusing first visit. Every design decision covered here, from proxy access to account recovery to how a lab result gets worded, ultimately serves that one number. A portal that’s technically compliant and quietly unused hasn’t solved the problem it was built for.
Tracking that number requires more than a login count in an analytics dashboard. It means watching what happens after a patient logs in, whether they finish the task they came for or call the front desk anyway a few minutes later. It means noticing whether they come back the next time they have a question instead of reaching for the phone. A portal earns trust one completed task at a time, and losing sight of that in favor of a feature checklist is how a technically solid product ends up quietly unused anyway.
Frequently asked questions
Why do so many patients never log into a portal they’ve been given access to?
Usually friction, not disinterest. Confusing account setup, unclear consent language, or a recovery process that takes multiple phone calls all push patients back to calling the front desk instead of trying the app again. The gap between patients offered access and patients who use it is a design signal, not a patient-behavior problem.
Is HIPAA compliance mainly a backend concern?
No. Encryption and access logging happen behind the scenes, but consent screens, session handling, and account recovery are decisions patients directly experience. A portal can be technically compliant and still confuse patients about what they’re agreeing to share.
How should a portal handle proxy or caregiver access?
As a core flow designed from the start, not a permission added after the single-user version ships. Caregiver access has grown quickly since 2020, and a portal that treats it as an edge case usually produces confusing consent screens that leave both the patient and the caregiver unsure what’s being shared.
Should a portal be built on top of the existing EHR or replaced entirely?
It depends on how locked in the organization is to its current EHR contract. A custom layer over the existing system can fix most patient-facing usability problems without a full rebuild, while a fully custom build makes sense for organizations with more flexibility and the resources to own compliance long term.
Why does billing inside a portal generate so many support calls?
Because most portals display billing data the same confusing way it appears on paper, rather than translating it. Line items, insurance adjustments, and visit dates that don’t obviously connect to each other push patients to call rather than trust what they’re reading on screen.
Does the mobile app need the same features as the web portal?
It needs the same reliability for the features patients use most, particularly scheduling and secure messaging. Treating the mobile app as a stripped-down version of the web portal frustrates exactly the patients most likely to use their phone as their primary way in.
What should a buyer ask a vendor before starting a patient portal project?
Ask for a specific walkthrough of a proxy access flow and an account recovery flow they’ve shipped before, not a general description. Ask how they test comprehension with real patients, not just internal staff, and how pricing adjusts once early usability testing surfaces scope changes.