Reliability Of Food Allergy Information Presented By An Artificial Intelligence Bot Versus Clinical Guidelines (2023 ACAAI presentation)




From our latest research:

🤖🍽 Artificial Intelligence Gives Artificial Information: A Closer Look at ChatGPT's Response to Food Allergy Guidelines 📚✨

Ever wondered how reliable AI chatbots like ChatGPT are when it comes to medical information? 🤔 Well, wonder no more! We recently delved into this fascinating topic and here's what we found:

🔍 Objective: Our mission was clear – to evaluate the quality and reliability of medical information provided by ChatGPT regarding food allergies.

📊 Methods: We put ChatGPT to the test by querying five key domains of the AAAAI/ACAAI food allergy guidelines. Then, we measured the information provided against the DISCERN instrument – a tool designed to assess health information quality. Spoiler alert: the results were eye-opening!

📈 Results: ChatGPT's responses scored an average of 3.4/5 on the DISCERN scale, indicating fair to good quality information. Not too shabby, right? But here's the kicker – while there was agreement with guideline recommendations, a whopping 61% of the sources provided by ChatGPT turned out to be unreliable. Yikes! 😱

🧠 Conclusions: While ChatGPT's responses seemed to align with established guidelines, the reliability of its sources left much to be desired. From fabricated journal articles to expired web pages, it's clear that caution is warranted when relying solely on AI for medical information.

So, what's the takeaway? While AI can be a valuable tool for accessing information, it's crucial to cross-reference and verify sources, especially in critical areas like healthcare. Stay informed, stay vigilant, and together, we can navigate the digital landscape with confidence! 💪💬 #AIinHealthcare #FoodAllergyAwareness #StayInformed

References:

https://www.annallergy.org/article/S1081-1206(23)00647-6/fulltext