The Web is a starting point for gathering information on every topic imaginable, including health care and disease. Websites offering information about health issues are popular among the general public. For people with vision loss, easy access to health information over the Internet may be especially important in overcoming difficulties accessing standard print information and traveling to health appointments. To examine the use and accessibility of health care websites, the American Foundation for the Blind (AFB) conducted a national survey of individuals with vision loss and analyzed the accessibility of key health-related websites. As the Affordable Care Act (popularly known as "Obamacare") goes into effect this month, it's a good time to look at what we found.
Survey Results Related to Accessibility of Health-Related Websites
In 2012, AFB conducted a national telephone survey of 400 adults with vision loss that focused on use of and satisfaction with health-related websites. Survey participants were also asked about importance of features such as text size, color contrast, labeled images, and search functions, and about the tools or methods they use to make websites more accessible and usable.
For the survey, health-related websites were put into three categories: general health information websites, drugstore or medical supply websites, and health insurance websites. More than three quarters (78 percent) of survey participants use health-related websites. Of those who used the websites, 80 percent reported using general health websites, one-third (32 percent) indicated using drugstore or medical supply websites, and only one-fifth (19 percent) of health website users said they visit insurance websites. Furthermore, the survey revealed that 75 percent of respondents reported being either satisfied or somewhat satisfied with general health websites and drugstore or medical supply websites. Insurance websites had the lowest satisfaction rating with 62 percent of users satisfied.
Slightly less than half of survey participants (45 percent) said they used tools or techniques to make information on the Internet more accessible. Among those who do, about a third (34 percent) said they used settings adjustment and about one-quarter (27 percent) used screen magnifiers. Other tools included glasses, the assistance of another person, screen readers, large monitors, and magnifying glasses. There were no participants who mentioned the use of braille and only five respondents (3 percent) reported using screen readers. This last statistic reveals that the results are strongly skewed towards the problems that individuals with some usable vision will face using a website as opposed to an individual accessing the site with a screen reader. Either our sample consisted mostly of individuals with very mild vision loss, or many respondents did not use access technology for some other reason. The respondents had difficulty naming a tool or technology that could improve their experience, so we believe that lack of information about access technology is a major contributing factor to the inaccessibility of health websites.
Website Evaluations
The second phase of the study was to perform a technical evaluation of the most commonly visited health websites according to the survey. This review of each website was carried out using the Web Content Accessibility Guidelines (WCAG 2.0), published by the Web Accessibility Initiative (WAI) overseen by the World Wide Web Consortium (W3C). This review was intended to determine the types of accessibility barriers that were present on each of the websites. The latest available version of Mozilla Firefox (version 20) and NVDA (version 2012.3) were used for evaluation. These programs were chosen because they are free and open source and because they tend to comply with Web standards.
The following websites were tested:
- General Health Information
- WebMD
- Mayo Clinic
- Online Drugstores
- CVS
- Walgreens
- Health Insurance
- BlueCross BlueShield
- United Health Care
- Humana
- Aetna
- State Health Exchanges
- Mahealthcareconnector (Massachusetts)
- Avenue H (Utah)
What Makes a Website Accessible?
According to the WCAG 2.0 an accessible website should be "Perceivable," "Operable," "Understandable," and "Robust." We will quickly review what each of these means and point out major shortcomings that were noted on the evaluated websites.
Perceivability
A website should be universally perceivable to all of its visitors. This means that the important information that is available to one user should be available to another user whether they're using a PC or Mac running Firefox, Safari with a screen reader, a magnifier, or any other compliant browser/operating system combination (called a user agent). This is specifically important to users who don't use vision to access a website but, instead, rely on a screen reader to interpret the underlying HTML. Problems arise when the Web developer, who uses his or her vision to access websites, creates a page that looks good visually but doesn't follow all design standards. Here is a representative list of some of the major issues that websites have with perceivability:
- No alternative text for non-text elements. It's obvious that a screen reader will not be able to interpret an image without alternative text, but there are other less obvious (to non-screen reader users) elements that need to be tagged as well. Input fields for forms is one example that causes major problems. A field can be "labelled" by visually placing text near a field, but if that label is not associated properly behind the scenes, the screen reader will not be able to make a connection between label and field. Videos and other time based media ideally should be transcribed and described for people who are hard of hearing or visually impaired.
- Low contrast text and other visual elements. The WCAG 2.0 guidelines have specific thresholds for different sizes of text and icons. In general, smaller text needs to be higher contrast, and there is a minimum contrast level for all visual elements no matter the size.
- Text within images. Websites are frequently designed with buttons, labels, headings, and other structural elements that include text in an image format, but this causes problems for both screen readers and screen magnification users. Oftentimes, these text images are not labeled, which is a problem for screen readers, but this convention also impacts screen magnification users because the image becomes pixelated when magnified while the text gets redrawn on the screen at high resolution.
The Perceivability standards are those that get violated most. Following the Perceivability guidelines of the WCAG 2.0 standards will fix most of the problems for screen reader and screen magnification.
Operability
This principle of the WCAG 2.0 guidelines focuses on the way that user agents interact with a website. Whether using a combination of keyboard and mouse, or only a keyboard or touchscreen, all ways of operating a computer should have the same access to a webpage. Here are a few common issues we found:
- Headings not used or used inappropriately. Headings are a very important navigation tool for screen reader users, and when implemented improperly (or not at all), it causes problems. There is a very specific set of guidelines for using headings, but in general there should be one top level heading, which tells you what the page is about, and, then, lower level nested headings used logically to indicate new parts of the webpage. Heading levels should be used sequentially and not be skipped.
- Multiple links with the same text but with different targets, e.g., "Click Here" links are a problem. The user has no idea what clicking there will do. Screen reader users frequently bring up a list of links on the page to make navigation easier. A list of "Click Here" links is not helpful.
- Keyboard traps. Sometimes a website will implement JavaScript (or another scripting technology), which moves keyboard focus around on the screen. Some implementations will actually trap the keyboard focus and not let you back on the main part of the screen. This is usually a bug, and it can require refreshing the page or leaving the page altogether to fix. This is annoying, and it also can make it impossible to complete certain tasks on the page.
- Illogical tab order. The focus should move in a logical fashion when a user tabs through the page. Sometimes the focus will jump across the page and skip over the content that is expected next.
- Mouseover events not occurring with the keyboard. When an element (like a drop down navigation menu) receives keyboard focus, the mouseover events should be registered so that the drop down actually occurs.
Forms, in particular, require a significant amount of interaction and can be a significant barrier to different users if not implemented using the Operability rules.
Understandability
A website can be perfectly perceivable and operable and still be inaccessible. A more human principle of the guidelines focuses on how understandable the content is to a person. In general, a website should use clear language, predictable navigation, and present the information that it's trying to convey in an understandable way. A few problems that we noticed on the websites surveyed are the following:
- Unpredictable navigation. A website should choose a convention and maintain it throughout the site. On some websites the navigation menu changes along with other conventions, such as calling the home page "Home" and, then, switching to "Main" on another page.
- Context changes upon interaction. A website with dynamic content should make it clear that the content will be changing. One example is a form that a user has to fill out that changes based on input from a field without informing the user. This can be confusing, especially for a screen reader user who will only become aware of the change after going back through the form.
- Poorly implemented input assistance. Sometimes a field will try to guess what you are entering and automatically complete the word. This is good if implemented properly but can cause confusion if abused. Auto complete and search-while-typing features can be cumbersome and problematic for screen reader users.
Robust
This principle states that the experience should be consistent for users across browsers and operating systems. The main point for this principle is that the HTML behind the webpage needs to conform to the HTML standards. There are some additional rules for how the browser should interact with technologies, such as Adobe Flash.
Summary of Results: Healthcare Website Accessibility Evaluations
The level of accessibility varied significantly across the websites that were evaluated. None of the websites evaluated could be considered fully accessible to screen reader users, and many of the sites do not work well for screen magnification users as well. Each site was tested with a markup validator. Essentially, this program looks over the code that makes up a website, and points out any inconsistencies with the standards. None of the websites we tested passed this validation, which is usually a strong indicator of poor accessibility. Five of the ten websites we looked at use the new HTML 5 standard.
Of the websites we evaluated, the Mayo Clinic, Humana, and Avenue H websites are the most accessible for both screen magnification and screen reader users. There are issues with each of the sites, but the solutions to these issues are simple to implement. There are no major issues that rendered these sites unusable.
CVS.com, Walgreens, mahealthconnector.org, and WebMD were mostly accessible but had some significant barriers to accessibility for some of the features of the sites. The major problem with the drugstore sites is their level of complexity. The main page, though technically accessible, is cluttered with so much information that it's very difficult to wade through. In addition, product listings in the store pages frequently had no alternative text or had repetitive links to the same information. WebMD is accessible in most places of the site except for the symptom checker tool, which is implemented in Flash and therefore is completely inaccessible to screen reader users. Finally, mahealthconnector.org is easy to use with a screen reader, but the use of text images for navigation links on the page make it difficult to use with screen magnification software. The images are labeled so that they are accessible to screen readers, but images of text should always be avoided because the quality of text in an image significantly degrades for screen magnification users.
BlueCross BlueShield, Aetna, and United Health Care all had significant accessibility issues with their websites at the time of evaluation.
Conclusions
As access to health information and healthcare is increasingly dependent on the Internet, it is critically important that these websites and information systems are made accessible to everyone regardless of the technology used to access them. Unfortunately, we have found that this is not the case thus far.
The level of accessibility of the surveyed sites varies significantly from site to site. Additionally, the level of accessibility that each site offers visually impaired users who access the content visually varies from the accessibility of screen reader users. The health information websites tend to have relatively good navigation and accessibility of text articles, but significant portions of those websites have inaccessible multimedia. The pharmacy store sites are technically accessible to both groups, but the complexities of the sites and the dynamic nature of storefronts make it difficult to navigate successfully. The health insurance websites we tested were poorly organized, poorly executed, and, in some cases, totally inaccessible to individuals trying to find information about health insurance. Since our initial tests, however, Humana, BlueCross BlueShield, Aetna, and United Health Care have all made significant changes that may have improved accessibility. We will reevaluate these websites and provide an update in a future issue of AccessWorld.
We'd like to hear about your experiences using health-related websites. We'd be especially interested in your experiences accessing state-based websites offering health insurance under the Affordable Care Act. Visit the federal government website on the Affordable Care Act and the availability of health insurance for more information. In future issues, we'll take a look at the accessibility of some of the state and federal government sites.
