Retinopathy of Prematurity (ROP) is the leading cause of preventable childhood blindness globally. As medical advances allow more premature and extremely premature infants to survive, the number of babies at risk of developing ROP continues to rise.
In Kenya, recent investments in maternal and neonatal health have led to establishment of newborn units even in remote and previously under-served regions. Yet while we have made important strides in saving premature infants, equivalent progress in protecting their sight is yet to be met.
With only about 150 ophthalmologists serving a population of 55 million, and most newborn units lacking specialised eye-care coverage, the majority of the 200,000 premature babies born annually do not receive the retinal screening they urgently need.
Local studies estimate that between 17 and 42 percent of Kenyan premature infants develop ROP. Many of these cases are preventable, and nearly all can be treated effectively when detected early.
Recognising this, Retinopathy of Prematurity Working Group of Kenya and Ministry of Health introduced national ROP guidelines in 2018, training for both trainee and practicing ophthalmologists, and strengthened screening programmes across the country.
A semi-informal but vital network of general ophthalmologists emerged, using binocular indirect ophthalmoscopy and smartphone-based fundoscopy to capture retinal images and share them remotely with paediatric ophthalmologists and vitreoretinal surgeons for guidance.
These innovations have improved early detection, lowered the severity of disease among diagnosed infants, and reduced the rates of visual impairment among those who require treatment.
Despite these gains, universal ROP screening remains out of reach. It is within this gap that telehealth and AI offer transformational solutions-solutions that Kenya has already begun to deploy. In March 2025, a tele-ophthalmology programme was launched at Kenyatta National Hospital’s newborn unit.
Today, KNH is now connected to newborn units at Mbagathi Hospital, Kiambu Level Four Hospital, Pumwani Maternity Hospital and Mama Lucy Kibaki Hospital. Trained technicians equipped with a mobile retinal camera rotate across these facilities each week.
For every eligible preterm baby, they capture six retinal images per eye, record clinical risk factors such as gestational age, birth weight, oxygen exposure and illnesses like sepsis, and upload the data for remote review by ophthalmologists.
Within just six months, the programme has screened 960 infants, of whom 360 were diagnosed with ROP, proving itself cost-effective and scalable. It offers a glimpse of a future where no child goes blind just because a specialist was hours away.
Telehealth, however, is only part of the solution with AI poised to revolutionise the landscape even further. Around the world, AI algorithms are increasingly being used to diagnose retinal diseases, from diabetic retinopathy to hypertensive retinopathy, with accuracy comparable or surpassing to that of human specialists.
Yet Africa risks being left behind, because most AI systems are trained using non-African datasets and may perform poorly when applied to African populations due to racial differences in anatomical features such as retinal pigment.
To correct this imbalance, Kenyan specialists are building an AI-ROP model specifically designed for African infants. Using retinal images gathered through the telehealth programme, the team is training a system that will be capable of detecting and staging ROP with high sensitivity.
Once fully developed, the AI tool will allow any healthcare worker to take retinal images using a smartphone and receive an instant preliminary diagnosis. This will dramatically reduce delays, ensure timely intervention and bring expert-level screening to remote newborn units that lack ophthalmologists.
Behind the statistics are real children and families impacted by untreated ROP; whose eyesight could be saved with simple, timely ROP screening. ROP-related blindness is almost entirely preventable, and Kenya now stands on the brink of becoming a continental leader in eliminating it.
To achieve this, we must expand tele-ROP services to counties nationwide, invest in training more technicians and equipping newborn units with retinal cameras, accelerate development and regulatory approval AI models and mobilize government and private-sector funding to sustain and scale these innovations.