Purchase this article with an account.
Ari Leshno, Emmanouil (Manos) Tsamis, Noga Harizman, Qing Wang, Sol La Bruna, Anvit Rai, Carlos G DeMoraes, George A Cioffi, Jeffrey M Liebmann, Donald C Hood; Assessment of Alternative Tests for the ICD-10 Glaucoma Severity Score. Invest. Ophthalmol. Vis. Sci. 2022;63(7):826.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The ICD-10 glaucoma severity classification is based on 24-2 visual field (VF) test. Given that 24-2 can miss central or early damage, we hypothesized that using the same principles, the optical coherence tomography (OCT) would provide a better estimation of severity.
Severity was determined for 55 eyes with glaucoma, according to the ICD-10 guidelines, which include assessing the presence of regional abnormalities in each hemifield, as well as the central 5 degrees of fixation (Fig. 1a). Grading was performed for each eye using either 24-2 VF or OCT macular and RNFL images by two independent groups of masked graders. Agreement between the two grading methods was evaluated. In addition, the accuracy of each method was determined by comparing the results to a reference standard (RS) based upon an automated structure-function topographic agreement method for the evaluation of regional abnormalities.
Based on the VF, glaucomatous eyes were classified as mild, moderate and advanced in 13, 17 and 25 cases respectively. Based on the OCT alone, glaucomatous eyes were classified as mild, moderate and advanced in 7, 4 and 44 cases respectively. The OCT classification was more severe than the VF in 22 (40%) of the 55 cases. In 26 (47%) cases, central involvement was detected by OCT, but not by the 24-2. In only 5 (9%) of the 55 cases, the OCT classification was less severe than the VF. The Kappa value showed weak agreement between VF and OCT (0.16±0.08, P=0.041). According to the RS, eyes were classified as mild, moderate and advanced in 3, 14 and 38 cases, respectively. The 24-2 gradings showed fair agreement with the RS (Kappa value 0.30±0.1, P=0.001), while the OCT gradings had good agreement with the RS (Kappa value 0.40±0.1, P<0.001). Compared to 24-2, the OCT also had better agreement with the RS in all 3 regions (Fig. 1b). A combination of using 24-2 to determine hemifield involvement and OCT for central involvement did not improve agreement compared to OCT alone (Kappa 0.36±0.1, P=0.001).
The current ICD-10 system relies on a functional (24-2 VF) test alone, which can underestimate the severity of glaucoma in a significant portion of eyes. OCT revealed advantages over 24-2 for accurate classification. Adding 24-2 to OCT data did not improve classification accuracy. 1. ICD-10 Glaucoma Reference Guide, 2019, aao.org; 2. Hood et al., PRER, 2013; 3. Tsamis et al., TVST, 2020.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.
This PDF is available to Subscribers Only