Stability of recorded sensitivities with the tablet-based Melbourne Rapid Fields (MRF) perimeter across eleven at-home weekly tests

Abstract

Purpose : Frequent visual field (VF) testing has been proposed to improve earlier detection of glaucoma progression and reduce visit burden for patients. Prospective observational clinical study. We evaluated variability of VF tests performed with tablet-based Melbourne Rapid Fields (MRF)

Methods : Subjects who were either glaucoma suspects or who had a glaucoma diagnosis were recruited. After a training session in the office, participants were loaned tablets loaded with MRF software and instructed to take VF tests at home daily for one week as a training period, then weekly thereafter for eleven weeks. If both eyes met eligibility criteria, participants were given the option to perform tests for both eyes. The 52 locations on the MRF 24-2 grid testing pattern that correspond to the Humphrey Field Analyzer (HFA) 24-2 Standard algorithm with normative values were analyzed. For each eye, slopes were calculated with sensitivity data for all locations to assess for stability over time.

Results : Seventy-three participants with a mean age of 61 ± 13 years (range 25 to 79) completed weekly MRF tests. Average recorded sensitivities from 1517 tests performed on 124 eyes were plotted. The locations with the highest variability were two peripheral spots next to the blind spot (Figure 1 and 2a). Location-specific slopes calculated from sensitivity data correspond to the weekly change in decibels (dB). When aggregated, the distribution of slopes for all 52 locations was centered on zero and whether slightly positive (n = 32) or slightly negative (n = 20), the slopes for 50 (96%) locations were within 2 SD from 0.0 (-0.056 to 0.056) (Figure 2b).

Conclusions : While individual point sensitivities varied from test to test, they were stable overall and reproducible in this population with relatively stable optic nerve status over a 3-month testing period. To detect a potential change in a visual field location, the sensitivity would have had to decline by 2 SD on average, with some points requiring a change of as high as 2.5 SD.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.

 

(a) 52 testing locations. (b) Mean threshold ± standard deviation across all tests. Locations 27 and 36 had highest variance

 

(a) Slopes calculated with sensitivities from 1517 tests performed on 124 eyes across 11 weeks. Distribution of all slopes was centered on 0. Locations 27 and 36 had the highest variance (boxed). (b) Average of slopes for each location