Tech-Enabled Health Equity: Bridging Language Barriers in Health Education

Healthwise Communications Team

Senior Asian female confused about online help with medications


One massive barrier to health literacy and ultimately health equity is also one of the most straightforward to solve: language. More specifically, communicating about people’s health in a language they understand.

The importance of multilanguage healthcare communication is clear: over 25 million people in the United States are considered to have limited English proficiency (LEP), and this is especially prevalent in Medicaid populations. And over 67.3 million people in the U.S. speak a language other than English at home. These linguistic barriers are proven to negatively affect healthcare access, patient experience, and quality outcomes.

It’s critical to address language barriers as part of larger health literacy and health equity strategies. However, engaging with patients, consumers, and caregivers in their preferred language can be costly and time-consuming—so it’s an investment that many healthcare organizations don’t make.

Healthwise is taking steps to unlock understanding by translating more of our health education into 18 languages beyond English and Spanish. And this is just the beginning!

Read on to hear from Alyson Erwin, Chief Product Officer, and Christy Calhoun, Chief Content Officer. They’ll share how expanding health information across languages will play a transformative role in achieving health equity and literacy goals, and how partnering with healthcare tech company Orbita to leverage generative AI and machine learning makes it possible on a large scale.

Q: Tell us a little bit about why Healthwise is making such a significant investment in expanding our multilanguage content offering. And why now?

Alyson: Providing multilanguage content is a way for us to further address the health literacy gap we see today and is in line with our mission to help people make better health decisions. By expanding our multilanguage offerings, we help our clients help their patients and members live healthier lives, by providing them with trusted, evidence-based content in their preferred language.

We’re doing this now because the ability for technology and generative AI to help us provide multilanguage content is finally at a point where we're confident the translations will be reliable, accurate, and consistent. Generative AI is an emerging technology that uses artificial intelligence, algorithms, and large language models to generate content. Machine translation, as a capability of AI, provides us with a way to expand our library quickly and easily but with a high degree of confidence in the results. Machine learning uses deep learning and neural network techniques to generate content based on the patterns it observes in a wide array of other content.

Christy: Healthwise aspires to be the trusted source of health education for all people. Expanding our offering of over 4,000 patient instructions to reach people in their own language is critical to our commitment to improving health literacy and helping address health disparities.

Q: We know that education translation can be time-intensive and expensive, so what makes this multilanguage expansion possible?

Alyson: The expanding capabilities of machine translation really make this possible for us. Translating content with full human support is very costly and time-consuming, but up to this point, we believed it was our best option for providing Healthwise content in other languages. So, we chose to translate a select portion of our content into 18 languages (in addition to English and Spanish) with human translators.

As technology has progressed, we feel confident that with our expert review, machine translation is the right choice for Healthwise and our clients. This technology, which improves daily, can provide accurate, reliable, and consistent translation for our Healthwise content. The Healthwise® Knowledgebase is our digital consumer health library that many of our clients know and love, and we’re expanding our full set of Knowledgebase content to 18 languages beyond English and Spanish. We’re also translating all 4,000 of our patient instructions into 5 additional languages, and we plan to further translate into 18 of the additional languages in the future. This increases the amount of multilanguage content we can provide our clients at a similar cost.

Christy: For these translation efforts, we have established a partnership with Orbita, an organization with a trusted reputation in training AI models and improving upon them in the healthcare space. We are excited to grow this partnership and expand our capabilities to do more together and serve more people with health education in their preferred language. Healthwise and Orbita share a commitment to improving the patient experience with clear, empathetic, and efficient communication.

Medical consultation between doctor and his patient.


Q: Why did we start with the languages we’re starting with?

Alyson: For our patient instructions, we took the 5 most-used languages from our current multilanguage patient instructions (Arabic, Brazilian Portuguese, Simplified Chinese, Haitian Creole, and Vietnamese) and expanded those languages across all our available patient instructions. We really wanted to provide a full set to clients who use our multilanguage content heavily.

For the Knowledgebase, we’re opting to use machine translation on demand through a chatbot. This enables us to easily add new languages so we can ultimately offer the full set of articles in 18 of the additional languages that we currently support.

Q: How do you see this particular expansion improving health outcomes for consumers? Can you talk about the link between health literacy and greater health equity?

Christy: Health literacy is an important driver of inclusive healthcare and better health. The more you empower someone with access to knowledge they can easily take in, remember, and act on, the greater the likelihood of impacting health choices and improving outcomes. A fundamental factor in addressing health equity is the need to bridge gaps through language access. By making education available in languages like Haitian Creole, Arabic, or Vietnamese, more people will have the opportunity to learn and take action. This can help people manage conditions like asthma to prevent unnecessary trips to the emergency department or choose to get vaccinations to keep themselves and their families and communities healthier.

Our new chatbot integration into the Healthwise Knowledgebase will aid in search and findability to get people directly to the health answers they want fast. That’s a benefit for speakers of English, as well as other languages. And the more we can help make learning quick and easy, the more we can help boost health literacy.

Q: How do you see this expansion improving the bottom line for healthcare organizations?

Alyson: One of the biggest challenges for healthcare organizations today is engaging people in their health journey. This is especially difficult for organizations whose patients either speak English as their second language or don’t speak English at all. By providing our clients with exponentially more translated content, we help them engage and educate the people they serve and eliminate their need for third-party translators or separate initiatives to complete content. Expanded multilanguage content ensures consumers are getting reliable, accurate health information they understand so they can take an active part in their health, which has been shown to lower overall costs.


Q: Why should clinicians and consumers trust education translated by a machine?

Alyson: Machine translation has made significant advancements in recent years, thanks to the development of sophisticated artificial intelligence models like neural machine translation (NMT). While machine translation is not yet perfect and can still produce errors and inaccuracies, it has reached a level where it can provide translations that are comparable in quality to manual translation in certain scenarios—including for health education. Here are a few reasons why machine translation is becoming as good as manual translation in some cases:

  • Neural networks and deep learning. Machine translation systems, particularly NMT models, utilize neural networks and deep learning algorithms. These models can process vast amounts of multilingual data, learn patterns, and make connections between languages. As a result, they can capture complex linguistic structures and produce translations that are more contextually accurate.
  • Large-scale training data. Machine translation models are trained on extensive datasets that include a wide range of texts from different domains and genres. These datasets allow the models to learn from diverse examples and improve their translation capabilities across various subjects.
  • Continuous improvement. Machine translation systems are continually being refined and updated. Developers and researchers use feedback loops, user evaluations, and comparison studies to identify and address the limitations of the models. Regular updates and improvements help bridge the gap between manual and machine translation.

Q: How does Healthwise ensure the medical accuracy and understandability of machine-translated content?

Alyson: This technology has come so far in recent years, we feel that we can provide this capability with limited review and it will still result in the high-quality, accurate, and reliable information you have come to expect from Healthwise. Before moving forward with the translation effort, we did a thorough quality check. We used the technology to translate some of our education into three different languages, and we not only had internal translators review it, but we also sent it to a third party for review. All reviews found the translations met our high standards for accuracy and plain language.

Get in touch with us to learn more about how offering non-English health education can help your organization address health disparities.