Global perspectives of ophthalmologists on artificial intelligence adoption in clinical practice

Abstract

Artificial intelligence (AI) is rapidly expanding in ophthalmology, yet its adoption in daily practice remains limited. Understanding clinicians’ perspectives is essential to address barriers and guide targeted education. We conducted a cross-sectional international survey of licensed ophthalmologists from 45 countries across all continents between October 2024 and February 2025. The questionnaire evaluated AI familiarity and use as primary outcomes, as well as perceived clinical impact, ethical concerns, and training preferences of participants. Descriptive and comparative analyses were conducted across world region, practice type, and professional seniority. A total of 622 ophthalmologists completed the survey. While 69.5% anticipated a moderate-to-very potential for AI to improve clinical outcomes, only 7.2% reported regular use. Familiarity with AI was significantly higher among academic clinicians (p = 0.0011), whereas 49.6% reported no knowledge of specific AI tools. Key barriers included lack of training (20.5%), implementation costs (16.5%), and reliability concerns (12.9%). Ethical issues most frequently cited were algorithmic bias (44.2%), liability (36.7%), and reduced physician–patient interaction (19.9%). Ophthalmologists with > 20 years of experience were more likely to support AI adoption (OR 1.5). Interest in AI education was high (75.1%), with a preference for online and structured formats and calls for earlier integration into medical curricula. Despite broad recognition of AI’s potential in ophthalmology, adoption remains low and familiarity limited. Lack of training, cost, and ethical concerns represent key barriers. Tailored, accessible education and institutional support are urgently needed to facilitate safe and effective AI integration into clinical practice.

Ähnliche Arbeiten