Glaucoma Detection in High Myopia with the Heidelberg Retina Tomograph 3

Abstract

Purpose: To compare the ability of the glaucoma probability score (GPS) and Moorfields regression analysis (MRA) using the Heidelberg Retina Tomograph (HRT) 3 to detect glaucoma in high myopia. Methods: Fifty-five eyes of highly myopic subjects (study group) and 139 eyes of non-highly myopic subjects (control group) were included. HRT images were obtained from all enrolled eyes, and agreement between GPS and MRA in all subjects, the study group, and the control group was established. Diagnostic performances of GPS and MRA were evaluated by including borderline classifications, either as test negatives (most specific criteria) or as test positives (least specific criteria). Results: The agreement between GPS and MRA in all enrolled eyes, the study group, and the control group was κ = 0.56, 0.50, and 0.58, respectively. In the study group, the AUC was 0.728 for GPS and 0.694 for MRA. In the control group, the AUC was 0.751 for GPS and 0.777 for MRA. With the most specific criteria, GPS showed 80.0% sensitivity and 60.0% specificity in the study group, and MRA showed 76.0% sensitivity and 63.3% specificity in the study group. With the least specific criteria, GPS showed 96.0% sensitivity and 40.0% specificity in the study group, and MRA showed 84.0% sensitivity and 40.0% specificity in the study group. Conclusions: In high myopia, GPS showed a high level of sensitivity with maintenance of specificity, compared with MRA using HRT 3. In addition, GPS had higher diagnostic performance than MRA in high myopia; however, there was no statistical difference between the AUC of GPS and MRA.