Mariam Hussain
Artificial intelligence (AI) is increasingly being integrated into ophthalmology, from automated screening tools, such as those for diabetic retinopathy, to newer AI models and research aimed at individualising patient care, patient education and improving surgical accuracy. Understanding patients’ attitudes is crucial for the successful integration of AI in ophthalmology; however, their perspective has seldom been studied. This article aims to explore this gap by reviewing the literature on patient-reported outcomes.
A major breakthrough in ophthalmology was the landmark approval of IDx-DR by the US Food and Drug Administration in 2018 for automated detection of referable diabetic retinopathy, marking the first widespread adoption of an AI system in any field of medicine (1).
In one of the largest studies to date evaluating the patient perspective on AI in ophthalmology, Yap et al (2) surveyed 438 patients undergoing diabetic retinal screening across New Zealand regarding their opinions on AI in ophthalmology. The survey consisted of 13 questions and covered topics such as awareness, trust, and receptivity toward AI systems. Interestingly, the authors observed that although 73% of participants were aware of AI, only 58-59% knew it could be used in healthcare or for diagnosis, highlighting the need for patient awareness. The survey also showed that patients would be more likely to trust AI if it is supervised by a clinician (2).
Similarly, in another study assessing patients’ perspectives on AI-assisted diabetic retinopathy, 121 patients completed a questionnaire about AI in diabetic retinopathy screening (3). Similar themes emerged, with only 14% of participants rating their knowledge of AI as good or expert; however, the majority felt positively about its use when adopted alongside doctors (3). Thus, suggesting that patients are receptive of AI as a facilitative tool to improve care delivery rather than as a replacement of clinicians.
Evidence also suggests that patients are more receptive of AI in treatment decisions when safeguards are in place, decisions are checked, and accuracy is prioritised. A study surveyed 181 patients with macular disease to evaluate the acceptability of AI-assisted retreatment decisions based on retinal scans. The study found that patients were generally accepting of AI, but the most important factors were accuracy, error rate, and whether the decision was checked. The participants did not have a preference regarding who made the decision between AI and a human, provided that the decision was accurate (4). Thus, highlighting that transparency and accuracy are crucial for building patient trust in AI.
AI also has a role in patient education. Macri and colleagues evaluated the acceptability of an AI-generated presenter in postoperative vitreoretinal surgery patient education such as face-down positioning (5). Although the sample size was small (n=15), the majority of the participants found the video useful, rating the quality as ‘excellent’ (5). Thus, highlighting that AI is not only acceptable in screening and retreatment decisions but also has a role in patient education and postoperative care.
From screening to treatment and patient education, the literature shows that AI use in ophthalmology is well-received by patients. However, the message is clear: AI should be introduced as an assistive technology under clinician’s oversight. Moreover, it is important that organisations are transparent about data use, errors, and responsibility to maintain patients’ trust. Ultimately, the value of AI should be measured by whether it improves patients’ experience, rather than solely by clinical efficiency.
References
1. U.S. Food and Drug Administration. FDA permits marketing of artificial intelligence-based device to detect certain diabetes-related eye problems [Internet]. 2018 Apr 11 [cited 2026 May 16]. Available from: https://www.prnewswire.com/news-releases/fda-permits-marketing-of-artificial-intelligence-based-device-to-detect-certain-diabetes-related-eye-problems-300628218.html.
2. Yap A, Wilkinson B, Chen E, Han L, Vaghefi E, Galloway C, et al. Patients perceptions of artificial intelligence in diabetic eye screening. Asia Pac J Ophthalmol (Phila). 2022 May 1;11(3):287–93. doi:10.1097/APO.0000000000000525 PMID: 35772087.
3. Malerbi FK, Ventura BM, Fischer M, Penha FM. Patients perceptions of artificial intelligence in a deep learning-assisted diabetic retinopathy screening event: a real-world assessment. J Diabetes Sci Technol. 2024 May 1;18(3):750–1. doi:10.1177/19322968241234378 PMID: 38404014.
4. Clinton S, Foss AJE, Bloomfield PS. The acceptability to patients with macular disease to have retreatment decisions being made by artificial intelligence. Eye Open 2026 2:4. doi:10.1038/s44440-025-00011-7
5. Macri CZ, Bacchi S, Wong W, Baranage D, Sivagurunathan PD, Chan WO. A pilot survey of patient perspectives on an artificial intelligence-generated presenter in a patient information video about face-down positioning after vitreoretinal surgery. Ophthalmic Res. 2024 Sep 23;67(1):567–72. doi:10.1159/000541530 PubMed PMID: 39312898.