PURPOSE: The determination of homogeneous subgroups of age-related macular degeneration (AMD) is necessary for clinical and genetic studies; therefore, the development of a simple, reproducible, and discriminating classification is essential. In this second part of our study (SPA-2), we evaluated a selected list of items for atrophic AMD based on color photographs of fundus, red-free frames, autofluorescence, fluorescein angiography, indocyanine angiography, and Spectral-Domain OCT.
METHODS: Ten items for atrophy were chosen from the literature and clinical experience. Twenty eyes of 20 patients with atrophic AMD were included. For each patient, the grid was completed by five independent, experienced readers from our reading center and by an expert. The Kappa coefficient was calculated for each item.
RESULTS: The greatest agreement between observers was found for the item "presence of atrophy" (Kappa=1). The worst concordance was recorded for the item "size of atrophy" (Kappa=-0.0286±0.0769 to 0.1813±0.0835).
CONCLUSION: The classification of atrophic AMD is complex and currently not very consensual, hence the need for a discriminant and reproducible classification grid. The evaluation of our grid for atrophic AMD shows satisfactory agreement between observers for the majority of the items. Some modifications are proposed to make it more discriminant and reproducible.