OBJECTIVES: Diabetic retinopathy (DR) is the most common microvascular complication of diabetes, with rising prevalence, resulting in a considerable economic burden that could be mitigated through early screening. However, the implementation of early detection using conventional screening methods is constrained. The emergence of artificial intelligence (AI) has enabled the development of automated methods that overcame those limitations, offering promising opportunities to facilitate DR screening.

METHODS: An investigation was conducted to estimate the economic burden of DR by determining costs associated with relevant examinations and treatments. Two budget impact models were developed over five years’ time horizon. The first model aimed to simulate the economic impact of DR early detection and the second one focused on the economic value of the automated methods compared to conventional methods for their annual implementation.

RESULTS: According to the study, the estimated economic burden of DR in Algeria amounts to approximately 25 820 077USD (an average of 35,26 USD per patient per year). The results of the modeling demonstrate a reduction in RD advanced patients and consequently, a significant cost decrease of approximately 42%, including both conventional and automated early screening methods. The estimation of using AI-based automated screening methods reveals a decline in screening costs, varying across scenarios (reduction of 6% in the worst-case scenario and 63% in the complete automated transition scenario). Moreover, automated methods address inherent issues in conventional approaches.

CONCLUSIONS: DR represents a significant burden. Without intervention, those expenditures will continue to escalate. Early screening for DR emerges as a promising solution, to preserve vision and to alleviate this economic burden. Through advancements in AI, automated methods have been developed, effectively addressing the challenges associated with fundus image analysis during this screening process, while also reducing the associated costs.