Abstract
Purpose :
To study the interobserver concordance between two resident doctors and two ophthalmologists specializing in retina regarding the detection of characteristics of diabetic macular edema (DME) in optical coherence tomography (OCT) in naïve patients using the LET classification.
Methods :
A retrospective consecutive analysis was carried out of patients evaluated in our service who presented DME for 1 year. OCTs were independently analyzed by two residents, on the one hand, as well as by two retinologists on the other hand, determining the characteristics of DME using the LET classification. The data were evaluated using a concordance analysis for each pair of observers.
Results :
A total of 95 eyes of 70 patients were analyzed, 67,1% being male and 32,9% female, with a mean age of 66,13 years +/-11,73 years. In the evaluation of the DME characteristics according to the LET classification, the kappa coefficient of interobserver concordance in the resident doctors was: Location (L) 0,36; Extension (E) 0,46; and Traction (T) 0,52. On the other hand, the values obtained among the retinologists observers were: Location (L) 0,709; Extension (E) 0,664; and Traction (T) 0,264. All the results were statistically significant.
Conclusions :
Classification of DME by OCT using the LET classification produces an acceptable degree of agreement between different observers, although there are notable differences when it is done by physicians in training and when it is done by retina specialists in clinical practice.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.