Purpose: To establish reproducibility of multifocal visual evoked potential (mfVEP) and traditional pattern-reversal VEP (tVEP) in normals and relapsing-remitting multiple sclerosis (RRMS).
Methods: mfVEP (60-sector dartboard) was recorded twice within a month in 40 normals and 40 RRMS patients [25 eyes with last optic neuritis (ON) ≥6 months, 34 non-ON]. mfVEP amplitude and latency (ms) were calculated as mean logSNR and median relative latency, respectively, for all 60 sectors (global) and 9 regions. tVEP was recorded (15', 60' and 120' checks) in subsets of 34 normals and 30 RRMS patients. tVEP N75-P100 amplitude (µV) and P100 latency (ms) were obtained. Reproducibility was evaluated using intraclass correlation coefficient (ICC) and test-retest variability (TRV). ICC ≥ 0.75 was considered good.
Results: ICCs for global and regional mfVEP were >0.80 in all groups. ICCs for tVEP were >0.75 for all except latency in ON (0.52-0.68). For mfVEP or tVEP, TRV for amplitude was similar among all groups; TRV for latency (ms) was larger in ON (5.3 for mfVEP global, 10.3 for 60' tVEP) compared with non-ON (3.1, 8.3) and normal (2.3, 5.7) (p < 0.05 for all). When tVEP was analyzed using similar methods as mfVEP (logSNR and relative latency), mfVEP global measures showed better ICC and TRV than tVEP in all groups.
Conclusions: mfVEP and tVEP showed good reproducibility in normals and RRMS. TRV for mfVEP latency was larger in ON than normal or non-ON. mfVEP global latency's TRV was about half the respective values for tVEP in all groups, due to averaging of multiple responses.