Test-Retest Variability in Visual Field Testing Using Frequency Doubling Technology

Abstract

To quantify the magnitude of test-retest variability (TRV) for normal subjects in serial visual fields (VF) using the frequency doubling technology (FDT) instrument. Twenty-one healthy adults, aged 23 to 60 years, underwent four serial FDT VF tests, using the full-threshold C-20 program of the Zeiss-Humphrey FDT analyzer, on one randomly chosen eye. The VF tests were spaced 2 to 4 days apart. All subjects performed two preliminary FDT tests in order to minimize any learning effect. Test-retest variability was calculated as the standard deviation of each location's sensitivity value across the four VF tests. Mean TRV (+/-SD) for the entire field was 2.44+/-1.32 dB. Mean TRV (+/-SD) for the superior, inferior, nasal, and temporal hemifields were 2.48+/-1.3, 2.40+/-1.4, 2.40+/-1.3, and 2.48+/-1.3 dB, respectively. Mean TRV (+/-SD) for the foveal location, the 4 central, and the 12 peripheral locations were 2.49+/-1.4, 2.16+/-1.2, and 2.54+/-1.4 dB, respectively. TRV was found to be rather uniform across the visual field of the commercially available FDT device, with only a mild, clinically insignificant, effect of both eccentricity and age on TRV. Variability in the FDT VF, for normal subjects, was found to be more uniform than that of both standard and short wavelength automated perimetry. In addition, a strong inverse correlation was found, in normal subjects, between the mean sensitivity and TRV.