Pattern shift visual evoked responses. Two hundred patients with optic neuritis and/or multiple sclerosis.

Fifty-one patients with clinically pure optic neuritis (ON) and 149 with possible, probable, or definite multiple sclerosis (MS) were tested with pattern shift visual evoked responses (PSVER) and compared with a group of 43 normal subjects. Attention was paid to response latency, intereye latency difference, as well as differences in amplitude or duration of the major positive peak (P100). Abnormal PSVER cannot be recorded from everyone with confirmed ON. Abnormal responses were recorded from 91% of all patients (including those with MS) who had a history of ON, 57% of all MS patients, and 36% of patients without a history of ON or an abnormal eye examination. Measurements of amplitude and duration proved to be of little value in this setting. Though abnormalities of PSVER are not "specific" for ON or MS, because they also result from other disease processes, they afford more reliable, quantitative documentation of abnormal conduction in visual pathways than any other clinical test.