Feasibility of integrating artificial intelligence–driven glaucoma screening into routine ophthalmic referral pathways in Zimbabwe

Abstract

Glaucoma constitutes a critical and growing public health crisis across Sub-Saharan Africa, where the prevalence of primary open-angle glaucoma is markedly higher than the global average and the rate of late or missed diagnosis is alarmingly elevated. In Zimbabwe, a structural shortage of trained ophthalmic personnel, compounded by inadequate diagnostic infrastructure, inequitable geographic distribution of eye care services, and low public awareness, has rendered conventional glaucoma screening pathways largely ineffective. Artificial intelligence, particularly deep learning–based fundus image analysis, presented a transformative opportunity to augment routine ophthalmic referral pathways by enabling automated, high-throughput, and cost-effective screening. This paper critically examined the feasibility of integrating AI-driven glaucoma screening into Zimbabwe's existing referral architecture, drawing upon the epidemiological burden of the disease, the demonstrated diagnostic performance of AI algorithms, the structural characteristics of the Zimbabwean eye health system, and the technological, ethical, and regulatory challenges attending such integration. The analysis concluded that while the technical capability of contemporary AI systems is sufficiently mature to support deployment, realising this potential in a resource-constrained setting demands concurrent investment in digital infrastructure, regulatory reform, workforce capacity building, and equity-centred design of automated triage systems.