Population impact and cost-effectiveness of artificial intelligence-based diabetic retinopathy screening in people living with diabetes in Australia: a cost effectiveness analysis

Authors:Wenyi Hu1,2, Sanil Joseph1,2, Rui Li3,4,5, Ekaterina Woods1,2, Jason Sun6, Mingwang Shen5, Catherine Lingxue Jan1,2, Zhuoting Zhu1,2, Mingguang He1,2,7,8, Lei Zhang1,3,4,9

Affiliations:

  1. Centre for Eye Research Australia, Royal Victorian Eye and Ear Hospital, East Melbourne, Australia
  2. Department of Surgery (Ophthalmology), The University of Melbourne, Melbourne, Australia
  3. Central Clinical School, Faculty of Medicine, Monash University, Melbourne, VIC, Australia
  4. Artificial Intelligence and Modelling in Epidemiology Program, Melbourne Sexual Health Centre, Alfred Health, Melbourne, VIC, Australia
  5. China-Australia Joint Research Center for Infectious Diseases, School of Public Health, Xi’an Jiaotong University Health Science Center, Xi’an, Shaanxi, 710061, PR China
  6. Eyetelligence Pty Ltd., Melbourne, Australia
  7. School of Optometry, The Hong Kong Polytechnic University, Hong Kong, China
  8. Research Centre for SHARP Vision, The Hong Kong Polytechnic University, Kowloon, Hong Kong SAR, China
  9. Clinical Medical Research Center, Children's Hospital of Nanjing Medical University, Nanjing, Jiangsu Province 210008, China

The following is a summary of a research paper by the authors. For more information, please see the published full-text journal article: https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(23)00564-3/fulltext

Integrating artificial intelligence (AI) into diabetic retinopathy screening could prevent tens of thousands of cases of blindness while delivering significant healthcare cost savings.
A modelling study, led by Prof Lei Zhang, A/Prof Lisa Zhu and Dr Wenyu Hu from both the School of Translational Medicine, Monash University and the Centre for Eye Research Australia (CERA), revealed that implementing AI-powered retinal scanning could prevent up to 40,000 cases of blindness from diabetic retinopathy over a 40-year period, while potentially reducing healthcare system costs by AU$615 million.

The technology works by capturing images of the retina and analysing them using AI systems trained on thousands of examples of both healthy eyes and those affected by diabetic retinopathy. What makes this approach particularly promising is its accessibility – the AI-powered cameras require minimal training to operate and are more cost-effective than traditional screening methods. This could transform how diabetic retinopathy is detected, especially in underserved communities. The portable technology could be deployed in general practice clinics and remote locations, making screening available to underserved communities where regular access to eye care specialists may be limited.

The research suggests that if more than 80% of the diabetic population could be screened, it would not only help detect unknown cases of diabetic retinopathy but could also identify previously undiagnosed cases of diabetes. While there would be some increased costs associated with treating more cases of the disease, the earlier intervention would prevent progression to blindness – ultimately reducing the burden on healthcare services.
AI screening could work effectively alongside clinical expertise to protect community vision health. This research points to a future where routine diabetic retinopathy screening could become a standard part of primary care, potentially revolutionising how we approach the prevention of diabetes-related vision loss.

Source: Unsplash