Comparative analysis of artificial intelligence platforms in generating Post-Operative instructions for endoscopic transnasal skull base surgery

Ahlam H Alamri ^1,2, Alya AlZabin ³, Nasir Magboul ⁴

⁰Department of Otolaryngology-Head and Neck Surgery, College of Medicine, King Saud University, Riyadh, Saudi Arabia. ahlam.alamri@hotmail.com.

European Archives of Oto-Rhino-Laryngology : Official Journal of the European Federation of Oto-Rhino-Laryngological Societies (eufos) : Affiliated with the German Society for Oto-Rhino-Laryngology - Head and Neck Surgery +

|

November 17, 2025

View abstract on PubMed

Summary

Artificial intelligence (AI) tools show varied performance in generating postoperative instructions for Endoscopic Transnasal Skull Base Surgery (ETSBS). While readability was similar, AI platforms differed significantly in patient understandability and actionability, requiring clinician oversight.

Area Of Science

Medical Informatics
Artificial Intelligence in Healthcare
Patient Education

Background

Postoperative care for complex surgeries like Endoscopic Transnasal Skull Base Surgery (ETSBS) requires clear patient instructions.
Patient comprehension of recovery guidelines is crucial for successful outcomes.
Artificial intelligence (AI) presents a potential avenue for generating patient education materials.

Purpose Of The Study

To compare the readability, understandability, and actionability of postoperative instructions for ETSBS generated by three AI platforms: ChatGPT, DeepSeek, and Gemini.
To evaluate the suitability of AI-generated content for patient education in a complex surgical context.

Main Methods

AI platforms were prompted to generate ETSBS postoperative instructions.
Readability was assessed using Flesch Kincaid Grade Level (FKGL) and Reading Ease (FKRE).
The Patient Education Materials Assessment Tool for printable materials (PEMAT-P) evaluated understandability and actionability.
Statistical analyses included Kruskal-Wallis tests and Pearson correlation coefficients.

Main Results

Readability scores (FKRE, FKGL) showed no significant differences across platforms, though Gemini had the highest FKRE and lowest FKGL.
ChatGPT and DeepSeek demonstrated significantly higher understandability scores compared to Gemini (p=0.005).
Actionability scores were highest for ChatGPT, followed by DeepSeek and Gemini, without statistical significance.

Conclusions

AI-generated postoperative instructions for ETSBS exhibit significant variability in understandability and actionability, despite similar readability.
Current AI platforms do not consistently meet optimal standards for patient education materials.
Clinician review and refinement of AI-generated content are essential before clinical implementation to ensure patient comprehension and adherence.

Keywords:

Artificial intelligence ChatGPT DeepSeek Gemini Machine learning Otolaryngology Post-operative instructions