Smart Solutions, Sharper Vision: Can AI Transform Diabetes Eye Care in India?

“I never realized how common blindness is among people with diabetes until the nurse explained why I needed an eye screening. Good that they offer it here, as seeking ophthalmic care for early diagnosis or confirmation of diabetic eye disease is not a common practice,” says Rajya Lakshmi, a diabetic patient awaiting her turn at the IDEA clinic in Hyderabad. In India, every fifth adult lives with diabetes,1 and with it, a silent threat to their sight. In the country’s crowded outpatient clinics, it’s not uncommon to come across patients with diabetic retinopathy, some already blind while others unaware that their vision loss is linked to their diabetes. By the time they seek care, it’s too late.

Diabetic Retinopathy (DR) is a complication of diabetes that damages the retina and may cause irreversible blindness. It is a common yet preventable cause of vision loss, if detected in time. However, timely detection is precisely what eludes the millions. DR affects over one-third of people with diabetes globally.2 India bears a significantly high burden of diabetic retinopathy, with 12.5 percent of Indian adults affected.3 Along with the lack of awareness, inadequate screening coverage compounds the problem.

In traditional settings, DR screening would be done by an ophthalmologist, and the services are confined mostly to tertiary care settings, accessible to few. Ophthalmologists’ availability is also limited (23,000 registered nationally).4 Combined with a fragmented healthcare landscape, weak infrastructure and persistent socio-economic disparities, this creates a quiet crisis of preventable blindness. For a diabetic patient, the first point-of-contact is generally the physician. If eye scan were available at the local physician’s clinic, early detection would be far more likely. However, given the rising burden of diabetes in India, this screening must be streamlined to ensure maximum coverage without adding to the physicians’ workload.

That’s precisely the promise of SMART (AI-enabled) DROP (Diabetic Retinopathy Outcomes and Pathways), a health system innovation, increasingly employed at physician’s clinics. Using machine learning algorithms, SMART DROP can flag early signs of DR. It is portable, easy-to-use and operable by non-specialist staff, allowing for eye screening to move beyond tertiary centres to local clinics. By decentralising care, integrating technology into routine clinical workflows, and enabling task-shifting, SMART DROP reimagines who gets screened, when and where. It has the potential to reduce diagnostic lags, cover more patients and save precious time while relieving the ophthalmologists’ burden.

An ICMR-funded study led by Max Institute of Healthcare Management at the Indian School of Business (ISB-MIHM), in collaboration with LV Prasad Eye Institute and Institute of Diabetes, Endocrinology and Adiposity (IDEA clinics), assesses the effectiveness of deploying SMART DROP systems for early detection of DR. The study compares the effectiveness of deploying AI-powered screening at physician clinics against routine DR screening in tertiary settings, with focus on patient coverage, referral accuracy and patient satisfaction and adherence to follow-up. The insights from the study can help build smarter pathways for DR management, where precision and practicality go hand in hand, making diabetic eye care not just better but also within reach for millions.

Indian healthcare system stands at the cusp of a digital overhaul. Along with accuracy and cost-effectiveness, it is essential to use AI-enabled innovations such as SMART DROP responsibly. The study puts ethics and informed consent into practice, setting a crucial precedent for dealing with health data. When embedded with ethical safeguards, such innovations can inspire trust and acceptability among users, making the system efficient and sustainable.

As evidence from the ISB-MIHM study emerges, it could help determine how AI-based screening systems such as SMART DROP can strengthen diabetic eye care ecosystems in India. If every physician’s clinic becomes a screening point, every patient gains the chance to protect their eyesight before it’s too late. AI can empower health systems to do more than just preserve the sight. It can build a robust health system with a sharper vision where every patient is seen. This matters because when our systems see better, so do our people.

References:

1. Agarwal, M., Rani, P. K., Raman, R., Narayanan, R., L., S., Virmani, A., Rajalakshmi, R., Chandrashekhar, S., Makkar, B. M., Agarwal, S., Palanivelu, M. S., Srinivasa, M. N., & Ramasamy, K. (2024). Diabetic retinopathy screening guidelines for physicians in India: Position statement by the Research Society for the Study of Diabetes in India (RSSDI) and the Vitreoretinal Society of India (VRSI)-2023. International Journal of Diabetes in Developing Countries, 44(1), 32–39. https://doi.org/10.1007/s13410-023-01296-z
2. Zegeye, A.F., Temachu, Y.Z. & Mekonnen, C.K. Prevalence and factors associated with Diabetes retinopathy among type 2 diabetic patients at Northwest Amhara Comprehensive Specialized Hospitals, Northwest Ethiopia 2021. BMC Ophthalmol 23, 9 (2023). https://doi.org/10.1186/s12886-022-02746-8
3. Raman, R., Vasconcelos, J. C., Rajalakshmi, R., Prevost, A. T., Ramasamy, K., Mohan, V., … & Kulkarni, S. (2022). Prevalence of diabetic retinopathy in India stratified by known and undiagnosed diabetes, urban–rural locations, and socioeconomic indices: results from the SMART India population-based cross-sectional screening study. The Lancet Global Health, 10(12), e1764-e1773.
4. Bali, J., Bali, O., Sahu, A., Boramani, J., Senthil, T., & Deori, N. (2022). State of the nation survey on cataract surgery in India. Indian journal of ophthalmology, 70(11), 3812-3817.