.png){kind=icon}
.png){kind=icon}
Return
Artificial intelligence in pediatric otorhinolaryngology: Assessing readability, understandability, and actionability of postoperative instructions
.png){kind=icon}
PRE
AI
A
H
L
Y
M
D
DOI：10.1016/j.amjoto.2026.104798.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Background: Despite the increasing use of artificial intelligence (AI) platforms in healthcare communication, their efficacy in producing patient-facing materials for pediatric surgeries is still unexplored. The readability, understandability, and actionability of postoperative instructions generated by three AI platforms (ChatGPT, Gemini, and DeepSeek) for common pediatric otorhinolaryngology (ORL) surgeries were compared in this study. Methods: Postoperative instructions were generated from each AI platform using a standardized prompt. Readability was assessed using the Flesch-Kincaid Reading Ease (FKRE) and Grade Level (FKGL). Understandability and actionability were evaluated using the Patient Education Materials Assessment Tool for Printable Materials (PEMAT-P). Comparative statistical analyses and Pearson correlation coefficients were calculated. Results: ChatGPT showed the highest readability with FKRE 64.57 (+7.20) and the lowest FKGL 7.40 (+1.54), significantly outperforming others in FKRE (p = 0.030). The Gemini had the highest understandability (83.20% + 1.56; p = 0.027), while the DeepSeek led in actionability (71.10% + 3.81; p = 0.140). No AI platform met all the adequacy thresholds (FKGL <6, FKRE >60, PEMAT-P > 80%) at one time. Across the procedures, the differences in the readability, understandability, and the actionability were not statistically significant (all p > 0.5). The correlation analysis showed that there is significant inverse relationships of FKRE with FKGL (r =-0.711, p = 0.032) and FKRE with understandability (r =-0.669, p = 0.049). The actionability was not significantly correlated with any metric. Conclusion: Although each AI platform shown strengths in some domains, none consistently fulfilled the criteria for optimal patient education. These findings underscore the necessity for enhanced AI training utilizing pediatric and caregiver-specific data to improve the quality of postoperative guidance in pediatric ORL care.
Keywords:
Artificial intelligence
Otolaryngology
ChatGPT
DeepSeek
Gemini
Machine learning
Post-operative instructions
Journal
IF:
1.7
Papers:
87
Citations:
5.3K
Organization
Cited Papers