Abstract
Purpose: :
To determine whether, in glaucoma, perimetrically ‘blind’ parts of the visual field can be omitted from subsequent tests without significant loss of information, thus reducing the test time.
Methods: :
In a prospective observational cohort study, 96 glaucoma patients performed at least eight reliable standard automated perimetry tests (Humphrey Field Analyzer 30-2 SITA fast). The probability of measuring a <0 dB sensitivity as a function of the number of preceding consecutive <0 dB sensitivities in the same test location was calculated. The effect on the mean deviation (MD) of assuming a <0 dB sensitivity based on the <0 dB history in the same test location was determined. The potential test-time reduction was estimated by assuming a time saving of 6.5 seconds per omitted blind test location.
Results: :
Three consecutive <0 dB sensitivities in the same test location yielded a probability of 97% to observe a <0 dB sensitivity in at least 2 of 4 consecutive follow-up fields. With the additional requirement of at least 10 test locations with <0 dB sensitivity in the visual field, this probability was 96% already after two consecutive <0 dB sensitivities. Assuming these locations to be blind in future fields without testing them again resulted in a median MD change of only 0.2 dB (90th percentile 0.6 dB). The estimated mean potential test-time reduction was 38%.
Conclusions: :
Omitting regions of no interest in advanced glaucoma yields large time savings without a clinically significant loss of accuracy in the determination of the MD.
Keywords: perimetry • visual fields • clinical (human) or epidemiologic studies: systems/equipment/techniques