Variability found in psychophysical testing despite efforts to control

The nature of psychophysical testing for visual field damage makes variability hard to control, according to a recent study.

Even after adjusting for 11 factors, researchers were able to reduce the magnitude of test-retest variability by only about a third. Severity of defect and between-subject variation exerted the largest effect on variability in repeated testing, they found.

Eytan Z. Blumenthal, MD, and colleagues at the University of California, San Diego, and elsewhere, conducted a prospective, comparative observational study to quantify factors affecting visual field test-retest variability. The study included 86 patients — 41 normal patients, 10 patients with suspected glaucoma and 35 patients with stable glaucoma.

The researchers created a multiple regression model that incorporated 13 factors affecting test-retest variability. This model was used to evaluate both standard visual fields and short-wavelength automated perimetry (SWAP), seeking to determine the extent of each factor's effect.

Mean test-retest variability for the standard visual fields was 1.28 ± 0.87 dB in normal eyes, 1.53 ± 1.04 dB in suspect eyes and 2.2 ± 1.79 dB in glaucoma eyes.

Using SWAP, mean test-retest variability was 1.87 ± 1.35 dB in normal eyes, 1.86 ± 1.24 dB in suspect eyes and 2.68 ± 1.85 dB glaucoma eyes.

The authors noted that 11 of the 13 factors evaluated in the study, when totaled together, accounted for less than one-third of the total causes of test-retest variability. These factors — severity, location, eccentricity, study group, diagnosis, superior vs. inferior hemifield, nasal vs. temporal hemifield, one-vs.-two thresholds, age, mean pupil size, and pupil size variability — accounted for 31% of test-retest variability seen in standard visual fields and less than 20% of the variability seen in SWAP.

Between-subject variation accounted for 8% of test-retest variability in standard visual fields and 13.5% of SWAP.

The authors noted that residual variation exhibited the greatest effect, accounting for 61% of the variability seen in standard visual fields and 66.6% in SWAP.

“This component seriously undermines our optimism in the ability to reduce the variability of various visual field tests and to fine-tune progression algorithms,” they reported in Ophthalmology.