Abstract
PURPOSE: Understanding test-retest variability (TRV) of mesopic microperimetry is critical for defining meaningful treatment effects in retinitis pigmentosa (RP) trials. This study uniquely evaluates intra- and intervisit TRV and coefficients of repeatability (CoRs) for microperimetry parameters in RP patients with varying best-corrected visual acuity (BCVA) levels. METHODS: In this single-centre prospective cohort study, RP patients were assessed on two visits, 14.0 days apart. Patients were grouped by BCVA: low (≤20/50 Snellen; ≥0.4 logMAR) or moderate (>20/50 Snellen; <0.4 logMAR). Using Bland-Altman analyses, the CoRs for intra- and intervisit variability were determined for pointwise (dB), mean (dB), and volume sensitivity (dB*deg(2)) on mesopic microperimetry. RESULTS: Intravisit CoRs for mean, volume, and pointwise sensitivity were 1.7 dB, 353.2 dB*deg(2), and 8.6 dB, respectively, in the low-BCVA group (n = 32), and 0.9 dB, 254.5 dB*deg(2), and 7.3 dB in the moderate-BCVA group (n = 15). Intervisit CoRs for mean, volume, and pointwise sensitivity were 2.4 dB, 355.2 dB*deg(2), and 10.2 dB in the low-BCVA group (n = 31). The moderate-BCVA group (n = 16) showed smaller CoRs of 1.6 dB, 386.8 dB*deg(2), and 7.7 dB for mean, volume, and pointwise sensitivity. BCVA and mean sensitivity, but not fixation stability, are predictors of TRV for volume sensitivity. CONCLUSIONS: Due to significant TRV, pointwise sensitivity is an unreliable endpoint for RP patients, irrespective of BCVA. Mean sensitivity is suitable as an endpoint when BCVA is relatively preserved. Volume sensitivity provides additional spatial information, and shows promise as a clinical endpoint for assessing macular sensitivity changes on mesopic microperimetry in patients with RP.