Barriers and Enablers Influencing the Implementation of Artificial Intelligence for Diabetic Retinopathy Screening in Clinical Practice: A Scoping Review

Area of Science:

• Ophthalmology
• Medical Informatics
• Public Health

Background:

• Diabetic retinopathy is a major cause of preventable blindness globally.
• Artificial intelligence (AI) is increasingly utilized in clinical settings.
• This review synthesizes literature on barriers and enablers for AI in diabetic retinopathy screening.

Purpose of the Study:

• To identify and summarize barriers and enablers for clinical applications of AI systems in diabetic retinopathy screening.
• To provide insights for effective implementation of AI technologies in eye care.

Main Methods:

• Systematic scoping review following PRISMA-ScR guidelines.
• Searches across multiple databases (MEDLINE, Embase, etc.) and grey literature.
• Independent review by two researchers, with consensus from a third; narrative data synthesis.

Main Results:

• 18 studies published 2018-2023 across 10 countries were selected from 3844 articles.
• Barriers and enablers were categorized into healthcare system, professional, user, and IT domains.
• Key themes included clinical efficiency (system), education (professional), patient outcomes (user), and diagnostic performance (IT).

Conclusions:

• Growing evidence supports AI efficacy for diabetic retinopathy screening.
• Understanding system, provider, consumer, and technological factors is crucial for AI adoption.
• Informed clinical strategies can facilitate sustainable and effective implementation of AI screening tools.