Comparative Analysis of Readability in Glaucoma Information Generated by AI: ChatGPT-5 vs. Gemini 2.5 Pro: Comparative Analysis of Readability in Glaucoma Information Generated by AI: ChatGPT-5 vs. Gemini 2.5 Pro

Resmiye Nur Okudan

doi:10.66288/actamedi.2026.81

Authors

Resmiye Nur Okudan SBU Antalya Research and Training Hospital

DOI:

https://doi.org/10.66288/actamedi.2026.81

Keywords:

Glaucoma, Artificial Intelligence, Readability Index, Large Language Models

Abstract

Background: Glaucoma is a leading cause of irreversible blindness worldwide, requiring effective patient education and communication for early detection and management. Recently, large language models (LLMs) such as ChatGPT-5 and Gemini 2.5 Pro have emerged as potential tools for providing medical information. However, the readability of AI-generated responses remains an important concern, particularly for patients with varying levels of health literacy.

Objective: This study aimed to evaluate and compare the readability of glaucoma-related responses generated by ChatGPT-5 and Gemini 2.5 Pro.

Methods: A total of 30 glaucoma-related questions, compiled and validated by three specialists, were presented to both AI models. The generated responses were analyzed using multiple readability indices, including Flesch Reading Ease (FRE), Flesch-Kincaid Grade Level (FKGL), Gunning Fog Index, Automated Readability Index (ARI), Coleman-Liau Index, SMOG Index, Linsear Write Formula, Dale-Chall Readability Score, and Spache Readability Formula. Statistical analysis was performed using SPSS, with significance set at p < 0.05.

Results: Both models produced responses within a similar readability range, generally corresponding to middle- to high-school reading levels. Gemini 2.5 Pro consistently generated slightly more readable text across most indices. A statistically significant difference was observed only in the Flesch-Kincaid Grade Level (p < 0.001), with Gemini producing lower grade-level text compared to ChatGPT-5. Other readability metrics showed no significant differences between the models.

Conclusion: Both ChatGPT-5 and Gemini 2.5 Pro are capable of generating understandable glaucoma-related information; however, Gemini demonstrates a modest advantage in readability. Despite this, the overall reading level may still be too high for individuals with limited health literacy. Further refinement of AI-generated medical content is necessary to improve accessibility and ensure effective patient communication.

References

1. European Glaucoma Society (2014) Terminology and guidelines for glaucoma, 4 edn. PubliComm, Savona

2. Crabb DP, Smith ND, Glen FC et al (2013) How does glaucoma look?: patient perception of visual field loss. Ophthalmology 120:1120–1126 DOI: https://doi.org/10.1016/j.ophtha.2012.11.043

3. Weih LM, Nanjan M, Mccarty CA et al (2001) Prevalence and predictors of open-angle glaucoma: results from the visual impairment project. Ophthalmology 108:1966–1972 DOI: https://doi.org/10.1016/S0161-6420(01)00799-0

4. Garway-Heath DF, Crabb DP, Bunce C et al (2015) Latanoprost for open-angle glaucoma (UKGTS): a randomised, multicentre, placebo-controlled trial. Lancet 385:1295–1304 DOI: https://doi.org/10.1016/S0140-6736(14)62111-5

5. Leske MC, Heijl A, Hussein M et al (2003) Factors for glaucoma progression and the effect of treatment: the early manifest glaucoma trial. Arch Ophthalmol 121:48–56 DOI: https://doi.org/10.1001/archopht.121.1.48

6. Leske MC, Heijl A, Hyman L et al (2007) Predictors of long-term progression in the early manifest glaucoma trial. Ophthalmology 114:1965–1972 DOI: https://doi.org/10.1016/j.ophtha.2007.03.016

7. Deutsche Ophthalmologische Gesellschaft (2020) Deutsche Ophthalmologische Gesellschaft (DOG), Berufsverband der Augenärzte Deutschlands e. V. BVA, Bewertung von Risikofaktoren für das Auftreten des Offenwinkelglaukoms. Leitlinie von DOG und BVA. Ophthalmologe https://doi.org/10.1007/s00347-020-01169-4 DOI: https://doi.org/10.1007/s00347-020-01169-4

8. Tham YC, Li X, Wong TY et al (2014) Global prevalence of glaucoma and projections of glaucoma burden through 2040: a systematic review and meta-analysis. Ophthalmology 121:2081–2090 DOI: https://doi.org/10.1016/j.ophtha.2014.05.013

9. Hohn R, Nickels S, Schuster AK et al (2018) Prevalence of glaucoma in Germany: results from the Gutenberg Health Study. Graefes Arch Clin Exp Ophthalmol 256:1695–1702 DOI: https://doi.org/10.1007/s00417-018-4011-z

10. Kapetanakis VV, Chan MP, Foster PJ et al (2016) Global variations and time trends in the prevalence of primary open angle glaucoma (POAG): a systematic review and meta-analysis. Br J Ophthalmol 100:86–93 DOI: https://doi.org/10.1136/bjophthalmol-2015-307223

11. Astrom S, Stenlund H, Linden C (2007) Incidence and prevalence of pseudoexfoliations and open-angle glaucoma in northern Sweden: II. Results after 21 years of follow-up. Acta Ophthalmol Scand 85:832–837 DOI: https://doi.org/10.1111/j.1600-0420.2007.00980.x

12. Cedrone C, Mancino R, Ricci F et al (2012) The 12-year incidence of glaucoma and glaucoma-related visual field loss in Italy: the Ponza eye study. J Glaucoma 21:1–6 DOI: https://doi.org/10.1097/IJG.0b013e3182027796

13. Czudowska MA, Ramdas WD, Wolfs RC et al (2010) Incidence of glaucomatous visual field loss: a ten-year follow-up from the Rotterdam Study. Ophthalmology 117:1705–1712 DOI: https://doi.org/10.1016/j.ophtha.2010.01.034

14. De Voogd S, Ikram MK, Wolfs RC et al (2005) Incidence of open-angle glaucoma in a general elderly population: the Rotterdam Study. Ophthalmology 112:1487–1493 DOI: https://doi.org/10.1016/j.ophtha.2005.04.018

15. Hitzl W, Stollinger M, Grabner G et al (2006) The Salzburg-Moorfields Collaborative Glaucoma Study: first results of the prevalence and 5‑year incidences in this prospective, population-based longitudinal study. Klin Monatsbl Augenheilkd 223:970–973. DOI: https://doi.org/10.1055/s-2006-927102

16. Rowlands, G. et al. A mismatch between population health literacy and the complexity of health information: an observational study. Br. J. Gen. Pract. 65, e379–e386, https://doi.org/10.3399/bjgp15X685285 (2015). DOI: https://doi.org/10.3399/bjgp15X685285

17. Liu, L., Qian, X., Chen, Z. & He, T. Health literacy and its effect on chronic disease prevention: evidence from China’s data. BMC Public Health 20, 690, https://doi.org/10.1186/s12889-020-08804-4 (2020). DOI: https://doi.org/10.1186/s12889-020-08804-4

18. Berkman, N. D., Sheridan, S. L., Donahue, K. E., Halpern, D. J. & Crotty, K. Low health literacy and health outcomes: an updated systematic review. Ann. Intern. Ways to

19. Sharma, N., Tridimas, A. & Fitzsimmons, P. R. A readability assessment of online stroke information. J. Stroke Cerebrovasc. Dis. 23, 1362–1367, https://doi.org/10.1016/j.jstrokecerebrovasdis.2013.11.017 (2014). DOI: https://doi.org/10.1016/j.jstrokecerebrovasdis.2013.11.017

20. Williams, J. The Skills for Life survey: A National Needs and Impact Survey of Literacy, Numeracy and ICT Skills (The Stationery Office, 2003).

21. Teravainen-Goff, A., Flynn, M., Riad, L., Cole, A. & Clark, C. Seldom-heard voices Adult literacy in the UK. Adult Literacy report https://cdn.literacytrust.org.uk/ media/documents/Adult_Literacy_2022_report_FINAL.pdf (2022).

22. Bostock, S. & Steptoe, A. Association between low functional health literacy andmortality in older adults: longitudinal cohort study. BMJ 344, e1602, https://doi.org/10.1136/bmj.e1602 (2012). DOI: https://doi.org/10.1136/bmj.e1602

23. Wang, L. W., Miller, M. J., Schmitt, M. R. & Wen, F. K. Assessing readability formula differences with written health information materials: application, results, and recommendations. Res Soc. Adm. Pharm. 9, 503–516, https://doi.org/10.1016/ j.sapharm.2012.05.009 (2013). DOI: https://doi.org/10.1016/j.sapharm.2012.05.009

24. McCray, A. T. Promoting health literacy. J. Am. Med Inf. Assoc. 12, 152–163, https://doi.org/10.1197/jamia.M1687 (2005). DOI: https://doi.org/10.1197/jamia.M1687

25. Docimo, S. Jr., Seeras, K., Acho, R., Pryor, A. & Spaniolas, K. Academic and community hernia center websites in the United States fail to meet healthcare literacy standards of readability. Hernia 26, 779–786, https://doi.org/10.1007/s10029-022-02584-z (2022). DOI: https://doi.org/10.1007/s10029-022-02584-z

26. Pandiya, A. Readability and comprehensibility of informed consent forms for clinical trials. Perspect. Clin. Res. 1, 98–100 (2010). DOI: https://doi.org/10.4103/2229-3485.71864

27. Swartz, E. N. The readability of paediatric patient information materials: are families satisfied with our handouts and brochures. Paediatr. Child Health 15,509–513, https://doi.org/10.1093/pch/15.8.509 (2010). DOI: https://doi.org/10.1093/pch/15.8.509

28. Williams, A. M., Muir, K. W. & Rosdahl, J. A. Readability of patient education materials in ophthalmology: a single-institution study and systematic review. BMC Ophthalmol. 16, 133, https://doi.org/10.1186/s12886-016-0315-0 (2016). DOI: https://doi.org/10.1186/s12886-016-0315-0

Comparative Analysis of Readability in Glaucoma Information Generated by AI: ChatGPT-5 vs. Gemini 2.5 Pro