Imagine receiving a new prescription after leaving the hospital. The label reads:
"Take one tablet orally twice daily with food. Avoid concurrent use of NSAIDs."
For many clinicians, this seems straightforward, but for someone reading English as a second language with limited health literacy, it may be difficult. What exactly is an NSAID? Does "orally" simply mean swallowing the pill? Does "concurrent" mean at the same time or on the same day?
Healthcare communication barriers are not just about language. They are also about understanding.
Understanding Low Health Literacy and Its Consequences
The average reading level for U.S. adults is seventh grade, while healthcare materials are recommended to be written at a fourth to fifth-grade reading level.
Approximately 80 million U.S. adults have limited health literacy, and 48% of people with low health literacy have completed college or hold a graduate degree. Low health literacy is not about education but reflects the gap between how health information is written and how people actually process it.
Patients with low health literacy have higher rates of medication errors, missed follow-ups, and poor chronic disease management. Improving health literacy could save Medicare alone $25.4 billion each year and prevent 993,000 hospital visits.
Why Translation Alone Isn't Enough
Translation solves one problem: language. It does not automatically solve understanding. Healthcare providers tell us that word-for-word translation doesn't address conceptual comprehension. They highlight the need to:
- simplify content before it's translated,
- preserve plain language and structure across languages,
- confirm understanding through teach-back, not just deliver words.
A recent global survey of 31,000 adults across 30 countries found that healthcare providers remained the most trusted source of health information globally (40.7%). When communication fails through jargon and complexity, trust erodes with it.
Using plain English isn't just the best approach. In many cases, it is required.
What Plain Language Actually Requires
Across federal guidelines from the Plain Writing Act, CMS, NIH, and AHRQ, four principles appear consistently:
- Start with the patient: frameworks open with a question. Who is your reader, and what do they need to do with this information?
- Use everyday language: keep sentences to 20 words or fewer and replace clinical vocabulary with the words patients already use. "Prior to" becomes "before." "Utilize" becomes "use." "Concurrent use is contraindicated" becomes "do not take these together."
- Write in active voice: make the actor and the action immediate. "Take this pill with food" instead of "This medication should be taken with food."
- Confirm understanding: a readability score is not confirmation of comprehension. The best way to know if a patient understands discharge instructions is to ask them to explain them back.
How AI is Improving Health Literacy
Understanding plain language principles is one thing. Applying them consistently is another. Clinicians work under significant time constraints, and patients receive different versions of the same message throughout their care journey.
AI has the potential to improve health literacy in several important ways:
- Real-time clarification matters. Many comprehension gaps happen during the encounter itself, not just after the visit. AI can support providers by making explanations clearer in the moment, especially for medication instructions, follow-up plans, warning signs, and care coordination.
- Health literacy is not just about reading level. A patient may be highly educated and still struggle with medical terms, medication instructions, discharge plans, or chronic disease management. AI can help convert clinical language into practical next steps: what this means, what you need to do, when to worry, and when to come back.
- Consistency across the care journey is a major advantage. Patients often receive verbal instructions, interpreter-supported conversations, after-visit summaries, portal messages, and printed education that are not always aligned. AI can help keep the same plain-language message consistent across interpretation, documentation, translation, and patient-facing materials.
- AI can support teach-back, not replace it. The goal is not just to deliver words, but to confirm understanding. AI can help structure patient communication around teach-back principles, prompting clearer explanations and helping identify when instructions may need to be repeated or simplified.
AI's greatest contribution to health literacy is not faster translation alone, but the ability to turn complex medical information into clear, actionable, language-concordant guidance across the entire care journey, from the clinical conversation to the after-visit summary, discharge instructions, reminders, and patient education materials.