Evaluating the European School for Advanced Studies in Ophthalmology classification for diabetic macular edema in clinical practice

Abstract

Purpose : To evaluate the utility of the European School for Advanced Studies in Ophthalmology classification for diabetic macular edema (DME) assessing the interobserver agreement for the staging using optical coherence tomography (OCT) in naïve patients. Secondly to assess the staging problems that may arise.

Methods : We performed a consecutive retrospective analysis of all DME naïve patients who came to our service for 1 year. Two specialist retinal evaluators analyzed independently the OCT images and determined the DME stage using the European School for Advanced Studies in Ophthalmology (ESASO) classification. An interobserver agreement analysis was performed. In cases where staging could not be performed, we evaluated and categorized the cause for no-staging and an interobserver agreement analysis was also performed.

Results : We analyzed 93 eyes of 65 patients (64.6% male and 35.4% female) with a mean age of 66.2 +/- 12.1 years old. Observer 1 staged 60 out of 93 eyes (64.5%). Observer 2 staged 61 out of 93 eyes (65.6%). The Kappa analysis of intra-observer concordance for the staging was 0.508.

Out of 93 cases, observers 1 and 2 could not stage 33(35.5%) and 34 (34.4%) cases respectively as the OCT features of the DME in these cases did not correspond to any of the options provided in the ESASO classification. These findings were statistically significant (p=0.000)

The causes for no-staging were edema with very low thickness that does not alter the ellipsoid zone (T0C1E0/E1), edema with low thickness that does not alter the ellipsoid zone with disorganization of the inner retinal layers (T1C1/2E0D1), edema with increased thickness without cysts (T1-3C0) and Edema with low thickness and severe cysts (T1C3). The Kappa analysis of intra-observer concordance for causes of no-staging was 0.869.

Conclusions : We may conclude that the EMD OCT ESASO classification constitutes a modern and thorough classification for the DME but in our experience does not englobe the totality of DME cases that you may find in clinical practice.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.