The association between race and longitudinal visual field variability

Abstract

Purpose : To investigate racial differences in the variability of longitudinal visual field testing in a “real-world” clinical population, and to estimate the impact of differences in variability on the time to detect visual field progression.

Methods : The study involved data from 3,551 eyes from 2,126 individuals of European descent (ED) and 1,342 eyes from 853 individuals of African descent (AD) extracted from the Duke Glaucoma Registry (DGR). Patients with 5 or more standard automated perimetry (SAP) tests and at least two years of follow up were included. Ordinary least squares linear regression was performed on the SAP mean deviation (MD) values for each eye over time. The standard deviation (SD) of the residuals was used as a measure of variability for each eye. The SD of the residuals was compared between the two racial groups using a Wilcoxon rank sum test. Computer simulations were then used to reconstruct “real-world” SAP MD trajectories based on rates of change and estimates of variability from the populations. The time required to detect progression using MD trend-based analysis was obtained under different assumptions about rate of change and baseline disease severity, assuming annual frequency of testing.

Results : Mean (SD) follow-up time was 9.1 (4.0) years for ED and 9.4 (4.2) years for AD. The mean (SD) number of visual fields was 7.7 (3.0) and 7.1 (2.6), and the corresponding average rates of change were -0.21 dB/year (95%CI:-0.23 to -0.19) and -0.22 dB/year (95%CI:-0.26 to -0.19), respectively (P=0.60). The mean SD of residuals was larger for AD (1.58 dB[95%CI:1.50-1.64]) than ED (1.29 dB[95%CI:1.27-1.33]; mean difference:-0.27, P<0.001). The increased variability resulted in an estimated delay in time to detect progression in AD compared to ED. For a scenario of a rate of change of -0.5 dB/year and -10dB MD at baseline there was a delay of 2.2 years (annual testing, 80% power).

Conclusions : These results confirm previous findings from a longitudinal cohort study (Gracitelli et al. JAMA Ophthalmol 2018), but this time on large scale data extracted from electronic medical records of a “real-world” clinical population. SAP tests in AD individuals had approximately 22% greater variability than those of ED, despite similar follow-up and number of tests. The increased variability would lead to a significant delay in time to detect progression. The reasons for racial differences in variability remain to be determined.

This is a 2020 ARVO Annual Meeting abstract.