Responses are moderately understandable, not readily actionable, and written at college reading level
By Lori Solomon HealthDay Reporter
MONDAY, Aug. 28, 2023 (HealthDay News) — Artificial intelligence (AI) chatbots generally produce accurate information for the top cancer-related web searches, according to a report published online Aug. 24 in JAMA Oncology.
Alexander Pan, from State University of New York Downstate Health Sciences University in Brooklyn, and colleagues assessed the quality of information and presence of misinformation about cancers generated by four AI chatbots. The analysis included 100 text responses to the five most commonly searched queries (Google Trends) related to the five most common cancers (skin, lung, breast, colorectal, and prostate).
The researchers found that the quality of text responses generated by the four AI chatbots was good (median DISCERN score of 5), and no misinformation was identified. Understandability was rated as moderate using the Patient Education Materials Assessment Tool (median understandability score, 66.7 percent), while actionability was poor (median actionability score, 20.0 percent). Based on the Flesch-Kincaid Grade Level score, the responses were written at the college level.
“These limitations suggest that AI chatbots should be used supplementarily and not as a primary source for medical information,” the authors write.
Abstract/Full Text (subscription or payment may be required)
Copyright © 2023 HealthDay. All rights reserved.
