Fluctuations on the Humphrey and Octopus perimeters.

Fluctuation of threshold determinations must be considered when evaluating visual field abnormalities. The short- and long-term threshold fluctuation were measured on 36 normal subjects using the Octopus and Humphrey automated perimeters. Subjects were tested using similar threshold programs that tested 16 locations twice. Short-term fluctuation was greater on the Octopus perimeter (1.6 +/- 0.3 dB) than on the Humphrey perimeter (1.3 +/- 0.3 dB; P less than 0.001). The heterogenous long-term fluctuation was greater on the Octopus perimeter (1.6 +/- 0.4 dB) than on the Humphrey perimeter (1.3 +/- 0.5 dB; P less than 0.05). The homogenous long-term fluctuation was near 0 for both machines. There are numerous variables that can alter the reproducibility of perimetry. Reproducibility can be estimated by short- and long-term fluctuation, but differences between the Octopus and Humphrey also influence their fluctuation.