- OT
- View all news
- Improving access to sight loss certification among glaucoma patients
Improving access to sight loss certification among glaucoma patients
UK and Australian clinicians have described the use of semi-automated screening technology that assesses CVI eligibility
13 May 2025
Researchers have described the performance of a semi-automated algorithm that screens glaucoma patients for certificate of visual impairment (CVI) eligibility.
Writing in British Journal of Opthalmology, clinicians described the use of the Glaucoma Field Defect Classifier (GFDC) to analyse the records of patients attending the glaucoma clinic at Cambridge University Hospitals NHS Foundation Trust between November 2020 and November 2021.
The tool used objective visual acuity and perimetry parameters to assess CVI eligibility.
The researchers found that 37% of glaucoma patients who were found to be eligible for CVI were not registered.
Reasons identified by the clinicians for this gap included administrative failure, frailty and comorbidity, reversibility of visual impairment, and mental health diagnoses.
Ophthalmology trainee, Dr Arun Thirunavukarasu, highlighted that he was one of 30 students who performed manual screening to identify glaucoma patients with sight loss.
“It became clear that this was not a practical or scalable solution for real-world clinical practice,” he said.
“We developed GFDC to enable more efficient screening, and the students ultimately used the app to identify blind glaucoma patients with much greater ease,” Thirunavukarasu highlighted.
Director of the Cambridge Eye Research Centre, Professor Rupert Bourne, extended his thanks to students at the University of Cambridge who contributed to the project.
“The team developed a semi-automated algorithm to identify glaucoma patients missing out on social care due to inaccurate, inefficient, and unreliable referral pathways that currently rely on ophthalmologists to recognise and implement certification of blindness,” he said.
Comments (0)
You must be logged in to join the discussion. Log in