• Raj Swaroop Lavadi, Ben Carnovale, Zayaan Tirmizi, Avi A Gajjar, KumarRohit PremRPDepartment of Neurological Surgery, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA., Manan Shah, D Kojo Hamilton, and Nitin Agarwal.
• Department of Neurological Surgery, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA.
• World Neurosurg. 2024 Nov 20.

Background And ObjectivesAtlasGPT represents an innovative generative pre-trained transformer (GPT), trained using neurosurgery literature. Its ability to contour its response according to the training level of the user is unique; however, whether its responses can be comprehended at each user's training level remains unknown. This study aimed to analyze the readability of responses provided by AtlasGPT.MethodsTen queries were presented to AtlasGPT across its four user profiles (i.e., surgeon, resident, medical student, patient). A readability analysis was performed using multiple instruments on Readability Studio. Readability scores of user-specific responses were compared using ANOVA testing and post-hoc pairwise t-tests with Bonferroni correction. P-value < 0.05 was considered to be significant.ResultsAcross the readability instruments that were leveraged, significant differences in reading ease were observed across all user profiles on comparisons to the patient (p < 0.005). Readability scores for the medical student profile tended to show greater reading ease than the surgeon and resident profiles; these differences, however, were not significant. The mean grade levels for patient responses across multiple instruments ranged from 8.8-11.51. Only one output via the New Dale-Chall assessment was written at the level of 5th-6th grade.ConclusionsAtlasGPT-generated content demonstrates readability variations according to the user profile selected; however, the readability of patient content still exceeds recommendations set by United States departmental agencies, necessitating a call to action.Copyright © 2024 The Author(s). Published by Elsevier Inc. All rights reserved.

6 keywords...

hide…