Purpose: This study aims to determine the intrasession repeatability and agreement using two methods to analyze fixation stability recorded during microperimetry in patients with macular disease.
Methods: Fifty patients with macular disease were enrolled prospectively to perform two consecutive microperimetry examinations on the Nidek microperimeter-1 (MP-1) using the same 1° red cross for fixation, Humphrey 10 to 2 grid of 68 test loci, 4 to 2 stair-case strategy, and 200 ms duration stimuli of Goldmann III size. Fixational eye movement was recorded during the two microperimetry examinations, and fixation stability was analyzed in two ways: (1) qualitatively through fixation stability grade (unstable, relatively unstable, and stable) and (2) quantitatively by calculating the logarithm of the bivariate contour ellipse area (logBCEA) that encompassed 68% of fixation points. Agreement and repeatability were assessed by unweighted kappa and the coefficient of repeatability.
Results: The mean (SD) visual acuity (VA) was 0.34 (0.32) logarithm of minimum angle of resolution. The mean (SD) duration of microperimetry was 13 (4) and 12 (4) min during the first and second measurements, respectively. During the first test, 10, 19, and 21 patients had unstable, relatively unstable, and stable fixation stability grading, respectively. Seventy percent (kappa: 0.53) of the fixation stability grading remained unchanged during repeat measurement. Of the remaining, 14% had a one-step decline and 16% had a one-step improvement in stability grading. The mean logBCEA were 4.28 and 4.24 in first and second tests, respectively (p = 0.43). The coefficient of repeatability for logBCEA was 0.61.
Conclusions: The logBCEA method of analysis for fixation stability assessment is preferable. If logBCEA derived from a microperimetry task is used as part of the core outcome set in a clinical trial, a change of >0.61 may be considered to exceed the limit of test-retest variability.