This trial was designed to determine if artificial intelligence (AI)-supported diabetic retinopathy (DR) screening improved referral uptake in Rwanda. The Rwanda Artificial Intelligence for Diabetic Retinopathy Screening (RAIDERS) study was an investigator-masked, parallel-group randomized controlled trial. Patients ≥ 18 years of age with known diabetes who required referral for DR based on AI interpretation. The RAIDERS study screened for DR using retinal imaging with AI interpretation implemented at 4 facilities from March 2021 through July 2021. Eligible participants were assigned randomly (1:1) to immediate feedback of AI grading (intervention) or communication of referral advice after human grading was completed 3 to 5 days after the initial screening (control). Difference between study groups in the rate of presentation for referral services within 30 days of being informed of the need for a referral visit. Of the 823 clinic patients who met inclusion criteria, 275 participants (33.4%) showed positive findings for referable DR based on AI screening and were randomized for inclusion in the trial. Study participants (mean age, 50.7 years; 58.2% women) were randomized to the intervention (n = 136 [49.5%]) or control (n = 139 [50.5%]) groups. No significant intergroup differences were found at baseline, and main outcome data were available for analyses for 100% of participants. Referral adherence was statistically significantly higher in the intervention group (70/136 [51.5%]) versus the control group (55/139 [39.6%]; P = 0.048), a 30.1% increase. Older age (odds ratio [OR], 1.04; 95% confidence interval [CI], 1.02–1.05; P < 0.0001), male sex (OR, 2.07; 95% CI, 1.22–3.51; P = 0.007), rural residence (OR, 1.79; 95% CI, 1.07–3.01; P = 0.027), and intervention group (OR, 1.74; 95% CI, 1.05–2.88; P = 0.031) were statistically significantly associated with acceptance of referral in multivariate analyses. Immediate feedback on referral status based on AI-supported screening was associated with statistically significantly higher referral adherence compared with delayed communications of results from human graders. These results provide evidence for an important benefit of AI screening in promoting adherence to prescribed treatment for diabetic eye care in sub-Saharan Africa.

The impact of unmet eye care needs in sub-Saharan Africa is compounded by barriers to accessing eye care, limited engagement with communities, a shortage of appropriately skilled health personnel, and inadequate support from health systems. The renewed focus on primary health care has led to support for greater integration of eye health into national health systems. The aim of this paper is to demonstrate available evidence of integration of eye health into primary health care in sub-Saharan Africa from a health systems strengthening perspective.

Recently there has been a great deal of new population based evidence on visual impairment generated in sub-Saharan Africa (SSA), thanks to the Rapid Assessment of Avoidable Blindness (RAAB) survey methodology. The survey provides information on the magnitude and causes of visual impairment for planning services and measuring their impact on eye health in administrative “districts” of 0.5–5 million people. The survey results describing the quantity and quality of cataract surgeries vary widely between study sites, often with no obvious explanation. The purpose of this study was to examine health system characteristics that may be associated with cataract surgical coverage and outcomes in SSA in order to better understand the determinants of reducing the burden of avoidable blindness due to cataract. This was a descriptive study using secondary and primary data. The outcome variables were collected from existing surveys. Data on potential district level predictor variables were collected through a semi-structured tool using routine data and key informants where appropriate. Once collected the data were coded and analysed using statistical methods including t-tests, ANOVA and the Kruskal-Wallis analysis of variance test. Higher cataract surgical coverage was positively associated with having at least one fixed surgical facility in the area; availability of a dedicated operating theatre; the number of surgeons per million population; and having an eye department manager in the facility. Variables that were associated with better outcomes included having biometry and having an eye department manager in the facility. There are a number of health system factors at the district level that seem to be associated with both cataract surgical coverage and post-operative visual acuity outcomes. This study highlights the needs for better indicators and tools by which to measure and monitor the performance of eye health systems at the district level. It is unlikely that epidemiological data alone is sufficient for planning eye health services within a district and health managers and study coordinators need to consider collecting supplementary information in order to ensure appropriate planning can take place.

With a global target set at reducing vision loss by 25% by the year 2019, sub-Saharan Africa with an estimated 4.8 million blind persons will require human resources for eye health (HReH) that need to be available, appropriately skilled, supported, and productive. Targets for HReH are useful for planning, monitoring, and resource mobilization, but they need to be updated and informed by evidence of effectiveness and efficiency. Supporting evidence should take into consideration (1) ever-changing disease-specific issues including the epidemiology, the complexity of diagnosis and treatment, and the technology needed for diagnosis and treatment of each condition; (2) the changing demands for vision-related services of an increasingly urbanized population; and (3) interconnected health system issues that affect productivity and quality. The existing targets for HReH and some of the existing strategies such as task shifting of cataract surgery and trichiasis surgery, as well as the scope of eye care interventions for primary eye care workers, will need to be re-evaluated and re-defined against such evidence or supported by new evidence.