Background: To determine the reproducibility of the multifocal visual evoked potentials (mfVEP) test results in a clinical glaucoma setting, and the factors that affect variability.
Methods: This was an observational case series study. The monocular mfVEP test, having a 58-sector, pattern-reversal dartboard array, was performed twice within 4 weeks in both eyes of 29 adult open-angle glaucoma (OAG) patients and suspects, using AccuMap Opera Software (ObjectiVision Pty Ltd, Sydney, Australia). The AccuMap severity index (ASI), the mean amplitude, and the individual amplitudes in each sector were compared between the 2 tests using intraclass correlations (ICCs). The effects of the severity of mfVEP field defects and signal-to-noise ratio (SNR) on the reproducibility of these variables were determined using the McNemar test and the Spearman rank correlation, respectively.
Results: The average ICCs of the ASI and amplitudes in the 2 tests were 0.84 and 0.87, respectively. Two sectors in the right eye and 6 in the left eye had significant differences between the 2 tests (Wilcoxon signed-rank p < 0.05). Larger differences were observed in patients having lower SNR (Spearman p = 0.022). Forty-six of the 58 eyes stayed within the same diagnosis category on repeating the test (i.e., within normal limits or outside normal limits).
Interpretation: Although there were some isolated examples of clinically significant differences on repeating the mfVEP test in our patients, our results suggest overall good repeat reliability. The variability of the test was higher in patients having high noise levels during the test.