Electronic filter effects on normal motor and sensory nerve conduction tests

Electronic filtering of the recorded signal is a significant aspect of all electro-diagnostic testing, but the filter type and frequency band are infrequently reported. Although the effects of different high- and low-frequency filters have been hypothesized, they have been the subject of little study under actual clinical conditions. This investigation showed clinically and statistically (p less than 0.01) significant alterations in both motor and sensory nerve evoked responses produced by modification of filter settings within the range routinely used for recording. As the low-frequency filter was varied from 3 to 20 Hz; large differences were seen in amplitude, area and duration of the compound muscle action potential. When the high-frequency filter was changed from 10 to 2 kHz, the mean amplitude of the sensory nerve action potential decreased by 12% and the mean latency was increased 0.1 ms. We conclude that filter parameters must remain constant when determining normal values and when performing serial studies on any patient. Filter settings should be reported as part of all electro-diagnostic reports and scientific manuscripts.