The Use of Artificial Intelligence Chat To Improve Readability of Patient-Facing Materials


Gauthier M1, Egan S2, Johnson N3
1Lumanity, Boston, MA, USA, 2Lumanity, San Francisco, CA, USA, 3Lumanity, Long Beach, CA, USA

OBJECTIVES: To examine artificial intelligence (AI) chat’s ability to assess and improve the readability level of patient-facing materials, inclusive of patient-reported outcome (PRO) measures.

METHODS: AI chat was used to evaluate the readability level (i.e. Flesch-Kincaid and Coleman-Liau formulas) of text from an established PRO measure, the SF-12, text from a patient interview guide, and an informed consent form (ICF). Depending on the established readability level, the chat was then queried to revise the text to be appropriate for a 6th or 3rd grade reading level. The appropriateness of recommended revisions were then evaluated.

RESULTS: The SF-12 instructions, item, and associated response options were determined to have a Flesch-Kincaid readability level of around the 10th grade and were revised to a 6th grade reading level by AI chat client. Similarly, interview guide text was revised from a 6th grade to a 3rd grade reading level, and the ICF was revised from a 7th grade to a 3rd grade reading level. AI-recommended revisions focused on simplifying the language, breaking down complex sentences, and removing complex terminology and/or replacing words with more appropriate vocabulary for each reading level. Revisions were determined to retain the intended meaning of the text. Some words (e.g., strenuous) were identified as candidates for further simplification.

CONCLUSIONS: AI chat can provide fast and effective insight into identifying challenging language in patient-facing materials and can inform revisions to improve readability. Revisions made by AI chat still need to be evaluated by researchers for appropriateness. AI chat cannot replace the value of engaging with patients directly, especially given the need to consider the unique perspectives of patients in various disease populations, including pediatric patients or patients with cognitive impairment. Including AI chat review of patient-facing materials and modifying materials accordingly is recommended prior to engaging with patients.

Conference/Value in Health Info

2024-05, ISPOR 2024, Atlanta, GA, USA

Value in Health, Volume 27, Issue 6, S1 (June 2024)




Methodological & Statistical Research

Topic Subcategory

Artificial Intelligence, Machine Learning, Predictive Analytics, PRO & Related Methods


No Additional Disease & Conditions/Specialized Treatment Areas

Your browser is out-of-date

ISPOR recommends that you update your browser for more security, speed and the best experience on Update my browser now