Visual Field Testing in a Telehealth Setting: Remote Perimetry Using a Head-Mounted Device in Normal Eyes

Abstract

Purpose : To determine the feasibility of remote visual field (VF) testing using a virtual reality visual field (VRVF) device with and without assistance.

Methods : Individuals without ocular disease (age ≥ 60) completed a standard automated perimetry (SAP) 24-2 SITA Standard on a Humphrey Field Analyzer and a VRVF (V₀) exam in random order on the same day in clinic with assistance. Participants were taught how to operate the VRVF equipment and given a device to take home.The following day, participants were telehealth-assisted to establish a remote testing environment, confirm proper use, and complete a VRVF exam (V₁). Participants tested remotely and without assistance once a week for 3 weeks (V_2-4). A pointwise mean absolute difference and SD between light sensitivity threshold values (LSTV) were calculated (V₀ vs SAP and V_1-4 vs V₀). Using a normal perimetric database for SAP ages 60-69, the number of points below normal (OD) was graphed. Data input was validated by a second party. Reliability indices were analyzed.

Results : 7 participants (14 eyes) were enrolled and completed reliable VRVF exams in clinic and remotely as determined by mean reliability indices (fixation losses ≤ 0.15, false positives ≤ 0.05, false negatives ≤ 0.06). The mean absolute difference and SD of LSTV between exams appear in Table 1. Excluding an outlier (P2 V₄) and missing data (P7 V₄), the absolute mean difference between V₀ and V_1-4 generally decreased as the study progressed. Figure 1 displays the number of LSTV below normal (OD), which decreased with weekly, remote self-testing (V_2-4) and was greater in those who took V₀ first (x_V0=14.5) than SAP first (x_V0=1.6).

Conclusions : Variability of LSTV from V₀ and number of points below normal reduced with weekly testing, implying that experience with testing improves results. Initial exposure to SAP may be useful for VRVF testing and could confound comparisons between V₀ and V_1-4. VRVF was generally reliable in a telehealth setting for individuals with normal eyes. Further research is needed to assess viability in a glaucoma patient population.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.

 

Table 1: Pointwise Mean Absolute Difference and Standard Deviation. Outlier P2 V₄ was excluded. P7 V₄ data was incomplete.

 

Figure 1: Number of Light Sensitivity Threshold Values (OD) Below Normal. Dotted lines indicate V₀ before SAP. Solid lines indicate V₀ after SAP. The outlier value (51) was not included in the mean.