RESUMEN
PURPOSE: To evaluate the ability of Stratus OCT and HRT II with Moorfields' analysis to detect localized RNFL defects and to evaluate the accordance of Stratus OCT and HRT II results in glaucoma patients. METHODS: A total of 60 patients (119 eyes) who had localized RNFL defects of either eye in red-free fundus photographs unserwent evaluation by Stratus OCT and HRT II. For the results of Stratus OCT and HRT II with Moorfields' analysis, normal distribution percentiles less than 5% were considered a significant RNFL defect. For each disc, superotemporal and inferotemporal portions were evaluated. The diagnostic abilities of Stratus OCT and HRT II to detect localized RNFL defects were calculated. The results of HRT II were compared with those of Stratus OCT. RESULTS: The overall sensitivity, specificity, positive predictive value and negative predictive value to detect localized RNFL defects were 67.9%, 88.9%, 84.3% and 76.5% in Stratus OCT and 67.3%, 64.8%, 62.2% and 69.6% in HRT II, respectively. Compared with Stratus OCT, 67.2% of HRT II results were in accord in the superotemporal portion and 68.9% in the inferotemporal portion. The accordance of detection of RNFL defects between the two instruments in the inferotemoral portion was higher (79.2%) than others. CONCLUSIONS: Stratus OCT with a normative database is a useful aid to detect localized RNFL defects in early glaucoma. If observation of topographic changes of the optic disc with HRT II is added, results will be even better.