How Many Subjects are Needed for a Visual Field Normative Database? A Comparison of Ground Truth and Bootstrapped Statistics

Purpose

The number of subjects needed to establish the normative limits for visual field (VF) testing is not known. Using bootstrap resampling, we determined whether the ground truth mean, distribution limits, and standard deviation (SD) could be approximated using different set size (x) levels, in order to provide guidance for the number of healthy subjects required to obtain robust VF normative data.

Methods

We analyzed the 500 Humphrey Field Analyzer (HFA) SITA-Standard results of 116 healthy subjects and 100 HFA full threshold results of 100 psychophysically experienced healthy subjects. These VFs were resampled (bootstrapped) to determine mean sensitivity, distribution limits (5th and 95th percentiles), and SD for different 'x' and numbers of resamples. We also used the VF results of 122 glaucoma patients to determine the performance of ground truth and bootstrapped results in identifying and quantifying VF defects.

Results

An x of 150 (for SITA-Standard) and 60 (for full threshold) produced bootstrapped descriptive statistics that were no longer different to the original distribution limits and SD. Removing outliers produced similar results. Differences between original and bootstrapped limits in detecting glaucomatous defects were minimized at x = 250.

Conclusions

Ground truth statistics of VF sensitivities could be approximated using set sizes that are significantly smaller than the original cohort. Outlier removal facilitates the use of Gaussian statistics and does not significantly affect the distribution limits.

Translational Relevance

We provide guidance for choosing the cohort size for different levels of error when performing normative comparisons with glaucoma patients.

A comparison of Goldmann III, V and spatially equated test stimuli in visual field testing: the importance of complete and partial spatial summation

PURPOSE

Goldmann size V (GV) test stimuli are less variable with a greater dynamic range and have been proposed for measuring contrast sensitivity instead of size III (GIII). Since GIII and GV operate within partial summation, we hypothesise that actual GV (aGV) thresholds could predict GIII (pGIII) thresholds, facilitating comparisons between actual GIII (aGIII) thresholds with pGIII thresholds derived from smaller GV variances. We test the suitability of GV for detecting visual field (VF) loss in patients with early glaucoma, and examine eccentricity-dependent effects of number and depth of defects. We also hypothesise that stimuli operating within complete spatial summation ('spatially equated stimuli') would detect more and deeper defects.

METHODS

Sixty normal subjects and 20 glaucoma patients underwent VF testing on the Humphrey Field Analyzer using GI-V sized stimuli on the 30-2 test grid in full threshold mode. Point-wise partial summation slope values were generated from GI-V thresholds, and we subsequently derived pGIII thresholds using aGV. Difference plots between actual GIII (aGIII) and pGIII thresholds were used to compare the amount of discordance. In glaucoma patients, the number of 'events' (points below the 95% lower limit of normal), defect depth and global indices were compared between stimuli.

RESULTS

90.5% of pGIII and aGIII points were within ±3 dB of each other in normal subjects. In the glaucoma cohort, there was less concordance (63.2% within ±3 dB), decreasing with increasing eccentricity. GIII found more defects compared to GV-derived thresholds, but only at outermost test locations. Greater defect depth was found using aGIII compared to aGV and pGIII, which increased with eccentricity. Global indices revealed more severe loss when using GIII compared to GV. Spatially equated stimuli detected the greatest number of 'events' and largest defect depth.

CONCLUSIONS

Whilst GV may be used to reliably predict GIII values in normal subjects, there was less concordance in glaucoma patients. Similarities in 'event' detection and defect depth in the central VF were consistent with the fact that GIII and GV operate within partial summation in this region. Eccentricity-dependent effects in 'events' and defect depth were congruent with changes in spatial summation across the VF and the increase in critical area with disease. The spatially equated test stimuli showed the greatest number of defective locations and larger sensitivity loss.