Evaluation of ChatGPT responses to common patient questions on ankle fusion
Mahmut Kalem, Hakan Kocaoglu, Peri Kindan Baltaci, Kamil Balaban, Ercan Sahin
article (Uluslararası, Hakemli) | Yazar(lar): Mahmut Kalem, Hakan Kocaoglu, Peri Kindan Baltaci, Kamil Balaban, Ercan Sahin | Yıl: 2025 | Dergi/Kurum: FOOT AND ANKLE SURGERY | İndeks: WOS.SCI | DOI: 10.1016/j.fas.2025.05.001 | Anahtar Kelimeler: ChatGPT, Artificial intelligence, Ankle fusion, Ankle arthrodesis, Ankle a
Background: The study assessed the quality and readability of responses given by ChatGPT to common patient questions about ankle fusion. Methods: Twenty-five frequently asked questions about ankle fusion were queried to ChatGPT 4.0 individually, and responses were assessed using the accuracy score developed by Mika et al. and the DISCERN tool. Readability was evaluated using the Flesch-Kincaid grade level, Gunning Fog, Coleman-Liau, and Simple Measure of Gobbledygook indexes. Results: ChatGPT's responses were generally of acceptable quality, with a mean accuracy score of 2, indicating that the overall responses were satisfactory and required minimal clarification, along with a DISCERN score of 49.78, which is considered fair. However, the readability level was high, with a mean of 11.6 grade level. Conclusions: ChatGPT showed promise as a resource for answering common patient questions about ankle fusion, providing mostly valid information. However, a high reading level was necessary to understand the response given. Levels of Evidence: N/A (c) 2025 European Foot and Ankle Society. Published by Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.