The effect of attention on conventional automated perimetry and luminance size threshold perimetry.

PURPOSE To investigate the effects of divided attention on conventional automated perimetry (CAP) and luminance size threshold perimetry (LSTP). METHODS Ten healthy subjects, ages 27 to 65, with two perimetry types (CAP and LSTP) were tested in random order. At a later session, these tests were given with a mental workload to simulate the effect of anxiety or distraction on subjects performing visual field testing, also in random order. The mental workload, the Paced Auditory Serial Addition Test (PASAT), was first administered to each subject, and the score was recorded. During the visual field testing, the PASAT was again administered continuously. Each subject was instructed to attend primarily to the PASAT while taking each visual field test. RESULTS CAP was affected by the addition of the PASAT, with a worsening of sensitivity from an average of 30.0 +/- 0.67 to 24.2 +/- 7.4 dB with a range of -0.04 to -23.2 dB (P = 0.04). LSTP showed a generalized reduction in threshold 1.71 +/- 0.22 to 2.35 +/- 0.72 dB with a range of 0.12 to -2.17 dB (P = 0.25). The percentage of correct responses on the PASAT was not significantly different between CAP (76.9%) and LSTP (74.8%). False-positive and -negative catch trial responses were increased during CAP with PASAT testing (P = 0.009). A substantial increase of fixation losses occurred during CAP with PASAT (3.7-16.2, P = 0.002). LSTP with PASAT showed increases in localization error (P < 0.001) and reaction time (P = 0.004). CONCLUSIONS Divided attention significantly affects performance on conventional automated perimetry with its fixed size stimuli and when the stimuli are scaled (LSTP). The deficits may simulate nerve-fiber-bundle-like defects.