Abstract
Purpose: :
To examine the impact of retinal field-of-view and magnification on the inter-expert reliability of plus disease diagnosis in retinopathy of prematurity (ROP).
Methods: :
15 wide-angle images from infants with ROP were cropped and adjusted in magnification to create two additional image categories: medium-angle (40-50°) and narrow-angle (20-30°). The resulting 45 images were uploaded to a web-based review system and interpreted independently by 13 ROP experts using a 3-level (plus, pre-plus, neither) and 2-level (plus, not plus) classification. Absolute agreement and kappa statistics were calculated to compare inter-expert reliability of plus disease diagnosis.
Results: :
In the 3-level classification, 70% or more experts agreed on the same diagnosis in 8/15 (53%) wide-angle images, but in only 3/15 (20%) medium-angle and 3/15 (20%) narrow-angle images. In the 2-level classification, 80% or more experts agreed on the same diagnosis in 11/15 (73%) wide-angle images, but only in 9/15 (60%) medium-angle and 3/15 (20%) narrow angle images. The mean kappa of each expert compared to all other experts was 0.40-0.59 in 8/13 (62%) experts using wide-angle images, was 0-0.19 in 7/13 (54%) experts using medium-angle images, and was 0.20-0.39 in 9/13 (69%) experts using narrow-angle images.
Conclusions: :
Inter-expert agreement in plus disease diagnosis from wide-angle images is higher than from medium-angle and narrow-angle images. Plus disease is defined using a narrow-angle standard published photograph which was selected over 20 years ago by expert consensus, yet this study suggests that peripheral retinal findings contribute to plus disease diagnosis by experts.
Keywords: retinopathy of prematurity