Can different types of retinal emboli be reliably differentiated from one another? An inter- and intraobserver agreement study

Objective: To determine whether ophthalmologists can agree on the qualitative assessment of visible retinal emboli. Design: Inter- and intraobserver agreement study. Setting: The retina and vitreous subspecialty session at the 1996 Canadian Ophthalmological Society meeting. Subjects: A total of 42 observers, of whom 30 were retinal specialists. Outcome measures: The observers viewed 17 fundus photographs of 11 patients with embolic acute retinal artery occlusion and classified the visible retinal emboli into one of three groups; cholesterol, calcific or other. Results: Overall, there was slight agreement for the 17 observations (mean κ = 0.063). The κ statistic for all cases ranged from slight to fair agreement. Slight interobserver agreement for the six unique photographs was observed (mean κ = 0.073). Slight intraobserver agreement was found for the three photographs that were shown in different orientations (mean κ = 0.041) and for the two photographs shown with differing magnification (mean κ = 0.102). Conclusions: Overall both intraobserver and interobserver agreement on the qualitative assessment of retinal emboli was poor. With only slight agreement on the classification of emboli, systemic, evaluation of acute retinal artery occlusion should not be based on qualitative assessment of retinal emboli.