<b>Cross-device consistency in diabetic retinopathy grading: a comparative analysis of widefield versus standard 30-35</b>° <b>- the ClarusDR study</b>.

Abstract

The Early Treatment Diabetic Retinopathy Study (ETDRS) 7-field protocol remains the reference standard for diabetic retinopathy (DR) severity grading, but is technically demanding, poorly tolerated and covers only 30-35% of the retina. Ultra-widefield (UWF) imaging systems, including Clarus and Optos, provide 200° retinal coverage with fewer acquisitions. This study evaluated agreement between ETDRS and UWF-based grading in non-proliferative DR. In this prospective study (NCT05746975), 81 type 2 diabetes participants (147 eyes) with ETDRS-Diabetic Retinopathy Severity Scale (DRSS) levels 35-47 underwent ETDRS 7-field colour fundus photography (CFP) and UWF imaging using Clarus 500 and Optos California. ETDRS grids were superimposed on UWF images. Two masked graders assigned DRSS levels using three strategies: ETDRS 7-field alone, periphery-only and extended (ETDRS plus periphery). Agreement was assessed and peripheral lesions were analysed for frequency and association with DR severity. Compared with ETDRS 7-field CFP, Clarus and Optos showed strong agreement in DR severity grading (Gwet's AC=0.774 and 0.809, respectively). Agreement was slightly lower for CFP-Clarus than CFP-Optos, reflecting higher detection sensitivity with Clarus, particularly for intraretinal microvascular abnormality and haemorrhages. Peripheral lesions were frequent and when combined with ETDRS fields identified eyes with more advanced disease. The peripheral retina alone does not show increased lesion severity, but extended UWF imaging integrating central and peripheral fields provides a more complete assessment. Extended Clarus and Optos imaging closely align with ETDRS grading while detecting additional severe cases, supporting UWF imaging as an alternative to ETDRS photography in clinical trials and practice. NCT05746975.