Variation in visual field measurements with an automated perimeter.

Using the Octopus automated perimeter, we conducted visual measurements on four separate occasions during a four-week period in six subjects with normal eyes. In 11 of 12 eyes, we found at least one point at which the threshold for light detection determined for that point showed a variance of more than 4 dB between tests. Thus, automated visual fields must be interpreted cautiously because an apparent pathologic change may only represent normal variation.