Abstract
Purpose :
To 1) assess the correlation between a new virtual reality visual field test, Vivid Vision Perimetry (VVP-10), and standard automated perimetry (SAP) in a global, sector-, and pointwise manner; 2) assess the structure-function relationship between VVP-10 and OCT; and 3) assess the test-retest variability of VVP-10
Methods :
Subjects with glaucomatous visual field defects were remotely trained to take the VVP-10 test and proceeded to take 10 tests over 14 days. VVP-10 response rate (fraction seen) at each point and SAP results including mean deviation (MD) and mean sensitivity (MS) at each point were recorded. Retinal nerve fiber layer thickness (RNFLT) at the optic nerve head was also recorded for each subject. Pearson correlation coefficients and the 95% confidence intervals were calculated for VVP-10 overall fraction seen vs SAP MD. Similar correlation coefficients were calculated for sector- and pointwise comparisons of VVP-10 average fraction seen vs SAP MS and structure-function analyses of RNFLT vs SAP MS and VVP-10 average fraction seen by sector. Test-retest variability was assessed with the intraclass correlation coefficient and a Bland-Altman plot of odd- vs even-numbered tests.
Results :
A total of 36 eyes from 19 subjects were included in the study. However, one subject who was unable to take the test properly and reproduce accurate results was identified as an outlier and excluded. The global correlation of VVP-10 overall fraction seen vs SAP MD was 0.90 (95% CI [0.80 – 0.96]). Sector-wise correlations between VVP-10 average fraction seen vs SAP MS ranged from 0.31 – 0.95; pointwise correlations ranged from 0.11 – 0.92. The correlation between RNFLT vs VVP-10 overall fraction seen was 0.49 [0.22 – 0.73] with sector-wise correlations ranging from 0.12 – 0.65. The correlation between RNFLT vs SAP MD was 0.57 [0.36 – 0.76], and sector-wise correlations of RNFLT vs SAP MS ranged from -0.14 – 0.62. The ICC was 0.97 [0.95 – 0.98] and one eye fell outside of the lower level of agreement in the Bland-Altman plot.
Conclusions :
Strong concordance exists between VVP-10 and SAP in a global, sector-, and pointwise manner. VVP-10 also demonstrates similar correlation with RNFLT compared to SAP. Finally, VVP-10 exhibits excellent test-retest variability, and its portable nature allows for remote testing and monitoring of glaucomatous field progression.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.