AI: The Next Opinion Leader in Health Communication

One of the critical challenges in health communication today is the sheer volume of health information available online. With over 70,000 health-related Google searches per minute¹, many people turn to the internet before consulting healthcare professionals. This trend is exacerbated by the widespread use of social media, where 60% of adults have searched for health information². Unfortunately, the majority do not verify the credentials of the sources, leading to widespread misinformation.

Misinformation in health is not only rampant but also dangerous. Over 40% of adults have encountered health misinformation online, with half of them unsure about its accuracy³. This uncertainty is compounded by the evolving nature of health evidence and the absence of consensus among healthcare professionals. We’ve all witnessed how contradicting information during the pandemic caused a lot of confusion and cost lives.

With this in mind, we consider that AI has continued to make significant strides in various fields, and has traditionally had a role in creating misinformation. But it also has an emerging role in combating misinformation, especially in health communication, as studied by Amelia Burke-Garcia, Director at the Center for Health Communication Science at NORC at the University of Chicago, and Rebecca Soskin Hicks, Pediatrician and Fellow at NORC. Together, they shed light on how AI can transform health communication and tackle the pervasive issue of misinformation.

The Role of AI in Misinformation

Misinformation has long been a serious issue in the medical field, especially with the emergence of social media. Rebecca, a practicing pediatrician, recalled a disturbing example from her clinical practice during the COVID-19 pandemic, where a patient was misled into using hazardous home treatments based on online misinformation. “This underscores the urgent need for accurate health communication”, she maintained.

AI, particularly in the dawning age of large language models (LLM), has the potential to both spread and combat misinformation. The rise of AI-generated content has made it easier to produce and disseminate false information. Studies have shown that large language models can generate vast amounts of misleading health content quickly⁴. For instance, researchers demonstrated that a handful of healthcare professionals could create hundreds of fake blog articles filled with vaccine misinformation in less than an hour⁵. AI’s ability to generate content is certainly impressive but it poses significant risks if not properly managed.

AI as a Solution

Despite these challenges, AI also holds promise as a tool to combat health misinformation. Through NORC, Amelia and Rebecca have been working on a new initiative, Health Communication AI⁶, which leverages AI to disseminate accurate and empathetic health information. Their vision involves developing an integrated AI system that uses social listening, sentiment analysis, and provides text responses through a Chatbot to identify and address misinformation in real-time.

Training these AI models requires high-quality, annotated data from authoritative sources. The goal is to create a chatbot that can engage with users empathetically and provide accurate health information. This involves not only technical solutions like prompt engineering to prevent bias but also participatory research to ensure the diverse needs of all communities are met.

The Future of Health Communication AI

The integration of AI into health communication is poised to transform how we address misinformation. AI-driven tools can act as opinion leaders, much like social media influencers, and other more traditional opinion leaders, providing credible and persuasive health information. Studies have shown that AI-generated responses can outperform human responses in terms of accuracy and empathy⁷, indicating a significant potential for AI to enhance health communication.

However, the deployment of such technologies must be done ethically and responsibly. This includes adhering to principles like non-maleficence, beneficence, autonomy, and justice. Ensuring that AI tools do not inadvertently harm users or perpetuate biases is crucial for their successful implementation.

Practical Tips for Identifying Health Misinformation

While AI continues to evolve, there are practical steps individuals can take to identify health misinformation today:

1. Consult Healthcare Professionals: Always verify health information with your doctor or a trusted healthcare provider.

2. Check the Source: Ensure that the information comes from credible and authoritative sources.

3. Evaluate Emotional Language: Be wary of content that elicits strong emotional responses or uses polarizing language.

4. Look for Miracle Claims: Skepticism is warranted for claims that seem too good to be true.

Conclusion

Developments in AI can be used for good as well as misinformation. Using AI to battle misinformation is both promising and complex. By leveraging the strengths of AI while addressing its ethical and practical challenges, we can enhance the accuracy and effectiveness of health communication. As we navigate this evolving landscape, the integration of AI into health communication strategies will be crucial in ensuring that accurate, empathetic, and effective health information reaches those who need it most.

For more insights and updates on this groundbreaking work, visit the NORC website and explore their Health Communication AI page.

References