I scanned 30 Hampton Roads clinic websites for accessibility. 90% had serious failures.

On November 3, 2026, I ran an automated accessibility scan on the websites of 30 independent healthcare practices across Hampton Roads. 27 of them (90%) had at least one failure the scanner rated serious or critical. Most of it comes down to low-contrast text and icon links that a screen reader can’t name, which are usually quick fixes.

Method

• Sample: 30 independent or small-group healthcare practices with websites, in Norfolk, Virginia Beach, Chesapeake, Portsmouth, Suffolk, Hampton, Newport News and Williamsburg. I picked practice types that commonly accept Medicare and/or Medicaid: family and primary care, combined primary and urgent care, pediatrics, OB/GYN, physical therapy, dermatology, behavioral health, podiatry and ENT (9 categories). I left out large hospital systems and national chains. A few of the included practices are larger independent groups with several locations.
• Attempted vs. counted: I tried to scan 35 sites. 5 could not be scanned and are not in any of the percentages below: 3 showed a bot-protection challenge that blocked the automated browser, and 2 returned a server error (HTTP 502) to the headless browser.
• Pages per site: the homepage plus one key patient-facing inner page, chosen from the homepage links in this order of preference: appointment request, new patients, patient forms, contact. 26 sites had 2 pages scanned. 4 sites had only the homepage scanned, because they had no separate page of that kind (they were mostly one-page sites). 56 pages in total.
• Tool: axe-core 4.10.2, run in headless Chromium (Playwright 1.63) at a 1280×900 desktop viewport, testing against the WCAG 2.0 and 2.1 Level A and AA rule sets (tags wcag2a, wcag2aa, wcag21a, wcag21aa). “Best practice” rules were not counted.
• Other checks: for each site I also recorded:
  • whether the scanned pages linked to an accessibility statement
  • whether an accessibility overlay script loaded (accessiBe, UserWay, AudioEye, EqualWeb and similar)
  • whether the new-patient or registration paperwork was offered only as PDF downloads. I reviewed each forms page by hand to separate patient intake forms from referral pads, job applications and the like.

Limits. This is an automated scan only. Automated tools reliably catch only part of the WCAG issues that exist on a page. They can’t judge whether alt text is meaningful, whether a form works with a keyboard and screen reader from start to finish, whether captions are accurate, or whether a PDF is tagged correctly. A site with “0 failures” here has passed the automated checks, which does not make it fully accessible. Each scan is also a snapshot of two pages on one day, on desktop. Element counts depend on page length and design, so they say more about how widespread a problem is than about how severe it is.

Headline numbers (n = 30 sites)

Every failure the scan found was rated serious or critical by axe-core, so “any failure” and “any serious or critical failure” come to the same 90%. Only 3 sites passed all of the automated checks.

Having an overlay or a statement did not mean a site passed. Both sites with an overlay widget, and 7 of the 8 with an accessibility statement, still had serious or critical failures.

Most common failures

Low contrast and unnamed links made up most of the problems. Low contrast alone accounted for about 500 of the roughly 720 affected elements (about 70%).

Breakdown by clinic category

Many categories have only a handful of sites, so read this table as a description of the sample, not as a ranking.

*Two other OB/GYN sites were attempted but could not be scanned (see Method).

What stood out:

• Dermatology and podiatry sites had the most kinds of failures. Missing alt text on images showed up there more than anywhere else.
• Behavioral health sites had the fewest kinds of failures (median 1). Even so, 4 of 5 had low-contrast text.
• ENT practices had the most PDF-only paperwork (2 of 3) and also the most accessibility statements (2 of 3).
• Family and primary care sites mostly had small numbers of issues. One site had more than 100 low-contrast elements across its two pages.

Takeaways for practice managers

• Start with contrast and links. Light-gray text, white text on photos, and icon-only buttons (phone, map, social media) without a text label caused most of the failures. A designer or web vendor can usually fix these fast, and the fixes help anyone reading on a phone in bright sunlight, not only people with low vision or people using screen readers.
• Check how patients get to you. The pages that matter most are the ones patients use to call, book an appointment, or find your office: phone links, appointment request forms, embedded maps and booking widgets. Have someone try to request an appointment using only the keyboard (Tab and Enter) to see where it breaks.
• Look at your paperwork. About 1 in 4 practices offered new-patient forms only as PDF downloads. A PDF that isn’t built for accessibility can be hard or impossible to fill in with a screen reader. Consider offering an accessible online form, or making sure the PDFs are properly tagged, and let patients know they can fill out paperwork with staff instead.
• A widget isn’t a fix. Both sites in this sample with an accessibility overlay widget still had serious failures in the underlying code. Changing the site itself is what makes it work for patients.
• Know the timeline. The updated HHS rule under Section 504 of the Rehabilitation Act sets WCAG 2.1 Level AA as the standard for websites of organizations that receive HHS funding, which can include practices that take Medicare Part A or Medicaid. Practices with 15 or more employees must comply by May 11, 2027, and those with fewer than 15 by May 10, 2028. Whether and how the rule applies to your practice is a question for your own counsel or compliance advisor, but those dates leave reasonable time to plan and budget for fixes.

No individual clinics are named in this report. Results come from an automated scan of public web pages on November 3, 2026, and are not a legal compliance determination.