LAUSR

Aims Certain patients with Diabetes Mellitus (DM) have high risk for complications from COVID-19. We aimed to test the hypothesis that pre-existing diabetic retinopathy (DR), a microvascular disease, is a prognostic indicator for poor COVID-19 outcome in this heterogeneous population. Methods Seven databases (including MEDLINE) and grey literature were searched, identifying eligible studies using predetermined selection criteria. The Quality in Prognosis Studies (QUIPS) tool was used for quality assessment, followed by narrative synthesis of included studies. Results Eight cohort studies were identified. Three showed significant positive associations between DR and poor COVID-19 outcomes. The highest quality study, McGurnaghan, found increased risk of the combined outcome fatal or critical care unit (CCU)-treated COVID-19 with referable-grade DR (OR 1·672, 95% CI 1·38–2·03). Indirectly, four studies reported positive associations with microvascular disease and poorer prognosis. Variability between studies limited comparability. Conclusions The current literature suggests an independent association between DR and poorer COVID-19 prognosis in patients with DM after controlling for key variables such as age. The use of standardised methodology in future studies would establish the predictive value of DR with greater confidence. Researchers should consider comparing the predictive value of DR and its severity, to other microvascular complications of DM.