Purchase this article with an account.
Michael Barry, Liancheng Yang, Rebecca Marcus, Gislin Dagnelie; Reproducibility of Goldmann Visual Field digitization by minimally trained operators. Invest. Ophthalmol. Vis. Sci. 2013;54(15):662.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To establish inter- and intra-operator variability among users of a manual digitization program designed to calculate areas within seeing and non-seeing isopters in Goldmann Visual Fields (GVFs) obtained in patients with inherited retinal degenerations.
GVFs collected in 25 patients with moderate to severe visual field loss were scanned and presented on-screen to 10 normally-sighted subjects. The subjects had attended a 1 hr training session explaining the meaning of GVF isopters and the use of the digitization software. During digitization, users mouse-clicked on 5 cardinal points in the GVF grid to establish the horizontal and vertical scales, and then identified each contour as seeing or non-seeing, selected the size of the test light (I-4e through V-4e), and digitized it by clicking on only those points along the contour that had been recorded by the GVF operators as transition points from non-seeing to seeing. The user completed all contours in a field before transitioning to the next field. Fields contained between 2 and 9 (mean 5.2) contours.
Each isopter of each test field was analyzed independently to determine the mean retinal area reported by the program, inter- and intra-operator variance of these data, variance as percent of the mean, and the number of operator errors per presentation. Data were subsequently averaged across presented isopters to estimate the expected variances and error rates of GVF digitization. The average inter-operator variance of retinal area calculations was 0.68 ± 1.17 (mm2 ± SD), or 2.38 ± 6.70 % of isopter area. Intra-operator variance was similar: 0.94 ± 1.78 mm2 or 1.47 ± 2.51 %. Operators had an average error rate of 1.06 ± 1.08 errors per isopter.
These results demonstrate that GVF digitization, with minimal training, produces consistent and repeatable retinal area measurements with reasonably small variance. In light of these data, GVF digitization can and will be used to track changes in visual fields of subjects with inherited retinal degenerations, ultimately to demonstrate the efficacy of treatments in clinical trials. Multiple operators and readers will be required to ensure proper labeling and completion of isopters in clinical trials. Data collection will continue to more accurately define the intrinsic variances of these methods.
This PDF is available to Subscribers Only