ABSTRACT
PURPOSE: To find the optimal parameter of retinal nerve fiber layer (RNFL) analysis using optical coherence tomography (OCT) and to compare the diagnostic precision between the new discriminant formula and the preprogrammed indicator (95%, 99% cut off value of normal RNFL thickness) in the diagnosis of glaucoma. METHODS: The study comprised 499 eyes of 284 patients. Routine ophthalmic examinations, fundus examination, visual field test, OCT optic nerve head analysis and OCT RNFL analysis were performed. Results were analyzed and classified into normal, suspected glaucoma, and glaucoma groups. RESULTS: There were 237 glaucomatous eyes and all parameters of the OCT RNFL analysis showed statistically significant differences between the glaucoma and non-glaucoma groups. The average RNFL thickness was most significantly correlated with the visual field indexes and was the most useful parameter in the detection of glaucoma. Using age, horizontal integrated rim area (HIRA) and average thickness, the accuracy, sensitivity and specificity of the new discriminant formula were 76.2%, 73.1% and 73.1% respectively. CONCLUSIONS: Average thickness was the most useful parameter in the detection of glaucoma. The next most useful parameter was inferior average thickness, and both were more precise than the preprogrammed indicator. The new discriminant formula is useful in the diagnosis and follow-up of the glaucoma patients.