Nearly 36% of adults in the United States have low health literacy skills. This means over a third of American adults will not fully understand their doctor’s instructions, much less the reasoning behind the instructions. This leads to worse adherence to treatment plans and therefore health outcomes. It is also part of the reason for skyrocketing healthcare costs.
These alarming facts underscore the critical need for better, more personalized health education. Gone are the days where we were limited to printed pamphlets with no room for edits or personalization. We live in a digital age, so let’s explore how generative AI can be used to transform the way we approach health education.
Take the case of Mary, a 55-year-old diagnosed with breast cancer. While Mary's oncologist spent time explaining her condition and treatment options, the medical jargon was overwhelming.
“I didn’t hear a thing after the words, ‘You have cancer.’”
This sentiment is almost universal for patients, regardless of existing health literacy capabilities. For Mary and millions just like her, traditional paper pamphlets offered little solace. She and her family stayed up late in the days after her diagnosis, searching for answers online and driving up their anxiety with the lack of clarity found online. This is where generative AI can make a difference.
Generative AI is a groundbreaking artificial intelligence technology that can create new, original content based on existing data. In the realm of healthcare, imagine a virtual tutor that could generate patient-specific educational material, tailored to their unique medical conditions, treatment plans, cultural background, and even their preferred learning style.
Imagine if Mary had received a personalized guide explaining her chemotherapy, adjusted for her underlying diabetic condition, and provided in her native language of Spanish. Generative AI can make this a reality by dynamically adapting to real-time changes in a patient's health profile.
Dynamic adaptation of educational materials can adjust to nuanced differences in a patient’s health literacy. Today's patient education materials usually comes in two or three different levels of literacy, geared toward 3rd or 5th grade reading levels. However, the reality is that patients’ literacy is far more diverse.
The ideal dynamic guide breaks down complicated procedures into digestible segments, using everyday language to explain complex medical terms. It does this specifically for Mary, rather than assuming she has the same educational needs as the next patient.
An even more advanced generative AI “pamphlet” might even allow Mary to query a specific phrase she doesn't understand, or “chat” with the content. This interactivity is non-existent in traditional education for obvious reasons; it would be near impossible to guess in advance all the possible questions patients would have about their health. With generative AI, this becomes possible, empowering the patient to interact and cement knowledge of their health so they can make their own treatment decisions with confidence.
Generative AI can even tailor information to be culturally sensitive and emotionally resonant. For example, generative AI-produced content might help a family of Asian decent to better understand their child's leukemia diagnosis by providing information in a culturally respectful manner, using familiar idioms and expressions.
Generative AI even excels at a surprising skill: empathy. Whereas doctors are typically running from one patient to another, leaving little time for softer skills to shine, AI content is always empathic in style, even when massive texts are generated on demand. To be clear, we are not advocating for technology to replace humans, but offering content in this manner could be the “human touch” that often gets lost due to lack of time in healthcare.
Generative AI doesn't come without risks. Ensuring data privacy and the medical accuracy of AI-generated content is paramount. Anyone who has tried to ask ChatGPT knows its annoying tendency to make up convincing references and so-called “real life” examples.
Using a human-in-the-loop system to validate the AI's recommendations is one way to ensure both the security and accuracy of educational material. There are also more technical solutions that capture origin data from a select set of documents or entire large language models fine-tuned specifically for the use case in question.
With nearly 40% of Americans struggling to understand basic health information, the need for personalized, accessible education has never been greater. Generative AI is poised to revolutionize the way we deliver health education, making it more personalized, accessible, and impactful than ever before.
.png)
