Global Health Expenditure Efficiency and Digital Gaps in Diabetic Retinopathy Disease Burden: A Panel Data Study of Inequalities (1990-2021)

Background Diabetic retinopathy is a leading cause of blindness globally, with significant disparities among countries in health expenditure efficiency and digital readiness. This study aimed to analyze the spatiotemporal distribution characteristics of diabetic retinopathy burden across 204 countries from 1990–2021 and explore associations with health expenditure structure, digital readiness, and ophthalmologist density. Methods Data were extracted from Global Burden of Disease Study 2021, WHO Global Health Expenditure Database, ophthalmologist density data, and Oxford AI Readiness Index. Panel data from 204 countries (2000–2021, 4,466 observations) were used to analyze associations between health expenditure and diabetic retinopathy burden. Threshold regression was performed on 177 countries to identify statistical breakpoints for AI readiness and ophthalmologist density. We acknowledge temporal misalignment in our data sources as a significant limitation, with AI readiness scores excluded from primary causal models due to temporal precedence issues. Findings Global age-standardized diabetic retinopathy YLD rates increased from 3.61 (95% UI : 2.39–5.21) per 100,000 person-years in 1990 to 5.45 (3.62–7.82) in 2021, representing a 51.1% increase. SDI showed an inverted U-shaped relationship with burden, with middle SDI countries having the highest rates (6.75 YLD/100,000). Each 1% increase in out-of-pocket expenditure proportion was associated with a 0.11% increase in diabetic retinopathy burden (p < 0.001). Due to temporal misalignment in data sources and the ecological study design, causal interpretations should be made with caution. Statistical breakpoints were observed at 67.68 ophthalmologists per million population ( R ²=0.1109, p  < 0.001) and AI readiness score of 56.01 ( R ²=0.0577, p  = 0.0044). Interpretation Middle SDI countries demonstrate a "development paradox" pattern in diabetic retinopathy burden distribution. Statistical associations suggest potential benefits from reducing out-of-pocket expenditures and achieving workforce density benchmarks, though these observational findings require prospective validation and contextual adaptation for policy implementation.