Follow one appointment request
Imagine someone trying to request an appointment through a health portal. They must find the right service, sign in, select a reason, choose a time, interpret any preparation instructions, and recognize that the request has been received. A problem at any step can prevent completion even when every page technically loads.
Now change the circumstances. The person might use a screen reader, have an unreliable connection, share a phone, or need instructions in another language. These are different circumstances with different design implications. Treating them as one generic category of “low engagement” would conceal the barriers that a service could actually address.
Accessibility, usability, and inclusion overlap
W3C's Web Accessibility Initiative distinguishes accessibility for people with disabilities from broader usability and inclusion concerns. The distinctions help keep specific needs visible. A visually simple page can still have inaccessible controls, while a technically accessible form can still ask confusing questions.
Usability evaluation asks whether people can accomplish a defined goal in the circumstances where they use the service. Inclusion brings additional questions about who can participate at all. A video appointment interface, for example, may be easy to operate on a stable connection without being a workable option for someone whose connection repeatedly drops.
These are not competing priorities. Clear labels, understandable feedback, and flexible interaction can serve several needs. But a design that works well for a few typical testers does not establish that it works for people using different assistive technologies or facing different barriers.
What a login count leaves out
A login records entry, not a completed task. A person might sign in repeatedly because they cannot find a result. Another might complete the same task once and never need to return. More activity is not inherently better; its meaning depends on the service's purpose.
Three possible meanings of repeated visits
- The service is useful for a recurring need, such as checking scheduled appointments.
- The person is waiting for information that has not yet appeared.
- The person is trying again after an unsuccessful attempt.
These hypothetical explanations call for different interpretations. A dashboard that reports only visits cannot distinguish them. Task completion, errors, time spent, requests for assistance, and feedback can make the pattern more understandable. Even completion needs a definition: submitting an appointment request is not the same as receiving a confirmed appointment.
Technical checks and user involvement answer different questions
Accessibility checks can identify issues such as missing programmatic labels or inadequate contrast. They are valuable, but no single automated score demonstrates that all users can complete every task. Some problems depend on interaction, content, and combinations of controls that require closer evaluation.
W3C guidance on involving users describes user evaluation as a complement to standards-based assessment. A person using their usual assistive technology may reveal obstacles that a checklist review misses. At the same time, a small user study cannot represent every disability or technology configuration. Combining evidence is more informative than treating either method as sufficient on its own.
Health information adds an interpretation task
A portal may make a laboratory result easy to find but difficult to understand. Unexplained abbreviations, unclear units, and vague instructions about what happens next can leave the person with access to data but little understanding of its significance. Providing more text is not automatically helpful if the important explanation becomes harder to locate.
The distinction between personal and organizational health literacy helps locate responsibility more fairly. A service controls many features of how information is presented and acted on. A confusing interaction does not establish that the person lacks interest or ability. It may reveal that the system assumed knowledge it never supplied.
Alternatives are part of the service
For some tasks, an alternative route can matter as much as the digital interface. A phone option that repeatedly redirects someone back to the inaccessible portal does not resolve the original barrier. Whether an alternative is workable depends on availability, response time, communication needs, and whether the person must repeat information.
This does not imply that every service can offer every possible channel. It means that access claims are more credible when they describe actual routes to completion and their limitations. A task-based evaluation can distinguish a convenient option from an effective service: who could finish, who could not, where the difficulty occurred, and what happened afterward. Those details make improvement more specific than asking people to become more engaged with a system that may not yet work for them.
