Abstract
Purpose: :
To evaluate the optic disk damage staging system (ODDSS); a new clinical method designed to classify optic disc (OD) appearance based on OD size and amount and localization of neural rim (NR) loss. Accuracy, reproducibility, reliability and clinical use were studied in normals and patients with open angle glaucoma (POAG). The ability of ODDSS in discriminating between normal and POAG eyes was compared to Heidelberg Retina Tomography (HRT).
Methods: :
This consecutive observational case study included 68 POAG patients (with reproducible glaucomatous visual field [VF] of various degree) and 40 healthy controls. ODDSS provides a clinical OD classification based on OD size (small, medium, large), severity of NR loss (6 stages), and localization of the NR loss (4 types). All subjects underwent SAP (30-2 SITA test), HRT II and clinical OD classification using ODDSS. ODs were graded with ODDSS in a masked fashion by 3 independent. ODDSS variability was also assessed in 30 eyes, in which OD grading was repeated in a similar manner within a one-month period. One eye per patient was considered for the analysis. Sensitivity and specificity in discriminating between normal and POAG eyes were determined for both ODDSS and HRT II Moorfield Regression Classification (MRC). Intra- and inter-observer variability in ODDSS scores and agreement with HRT II results were assessed using intraclass correlation coefficient (ICC) and Kappa test. The influence of VF defect severity (based on Glaucoma Staging System 2 [GSS 2] score) and OD size on the diagnostic accuracy of the ODDSS was assessed using logistic regression analysis.
Results: :
Sensitivity of ODDSS ranged between 73.5% and 89.7%; specificity ranged between 92.5% and 100%. Intra-observer ICC ranged from: .34 to .62 for OD size; .96 to .97 for damage severity; and, .56 to .58 for damage localization. Inter-observer ICC was respectively: .63 to .67; .93 to .97; and, .39 to .55. The agreement between ODDSS and HRT II MRC ranged from .66 to .93. ODDSS sensitivity in glaucoma detection increased with increasing VF defect severity (higher GSS 2 scores), which appeared more evident for larger ODs.
Conclusions: :
This preliminary study shows that ODDSS is an easy and applicable method in a clinical setting. ODDSS provides good accuracy, reproducibility and reliability compared to current objective methods like HRT.
Keywords: conjunctiva • tumors