M S Kook1, K Sung, R H Park, K R Kim, S T Kim, W Kang. 1. Department of Ophthalmology, Ulsan University School of Medicine, Asan Medical Center, 388-1 Pungnap-dong, Songpa-gu, Seoul, 138-736, Korea. mskook@www.amc.seoul.kr
Abstract
PURPOSE: To assess the reproducibility of measurement of retinal nerve fiber layer (RNFL) thickness among subjects using the GDx Nerve Fiber Analyzer (NFA). METHODS: One eye of each of five young and healthy volunteers was imaged by three newly trained technicians (R.-H.P, K.-R.K., and S.-T.K.). Each volunteer was scanned six consecutive times for five independent times within a 1-month period. One composite image was selected out of three best images per session. Intra- and interoperator reproducibilities were calculated and the influence of placement of measurement ellipse by a single operator on interobserver reproducibility was also investigated. RESULTS: Mean coefficients of variation for total integral value of RNFL were 3.67 +/- 1.47%, 3.86 +/- 2.13%, and 4.16 +/- 1.97%, respectively for the three technicians. Interoperator variability was not statistically significant (P = 0.075), and the difference was even less if a single ellipse drawn by one operator was used in all measurements (P = 0.1528) by repeated-measures analysis of variance (ANOVA). CONCLUSIONS: The results suggest that GDx NFA may provide acceptable intraoperator as well as interoperator reproducibility among Asian eyes.
PURPOSE: To assess the reproducibility of measurement of retinal nerve fiber layer (RNFL) thickness among subjects using the GDx Nerve Fiber Analyzer (NFA). METHODS: One eye of each of five young and healthy volunteers was imaged by three newly trained technicians (R.-H.P, K.-R.K., and S.-T.K.). Each volunteer was scanned six consecutive times for five independent times within a 1-month period. One composite image was selected out of three best images per session. Intra- and interoperator reproducibilities were calculated and the influence of placement of measurement ellipse by a single operator on interobserver reproducibility was also investigated. RESULTS: Mean coefficients of variation for total integral value of RNFL were 3.67 +/- 1.47%, 3.86 +/- 2.13%, and 4.16 +/- 1.97%, respectively for the three technicians. Interoperator variability was not statistically significant (P = 0.075), and the difference was even less if a single ellipse drawn by one operator was used in all measurements (P = 0.1528) by repeated-measures analysis of variance (ANOVA). CONCLUSIONS: The results suggest that GDx NFA may provide acceptable intraoperator as well as interoperator reproducibility among Asian eyes.
Authors: Robert Laemmer; Folkert K Horn; Arne Viestenz; Anselm G Juenemann; Christian Y Mardin Journal: Graefes Arch Clin Exp Ophthalmol Date: 2005-09-21 Impact factor: 3.117