Correlation analysis of somatosensory evoked potential waveforms: Clinical applications

Abstract

Objective: To assess the validity of the correlation coefficient (CC) as a diagnostic tool in evaluating somatosensory evoked potential (SEP) waveforms. Design: SEP waveforms were recorded over the spine and scalp in response to tibial nerve stimulation at the ankle for 30 patients. Setting: A general community, ambulatory, private referral center. Patients: Patients were all ambulatory and were selected randomly from among existing patient records at the medical center. Main Outcome Measure: The test for validity uses the CC to verify reproducibility. The test for side-to-side symmetry uses the CC to obtain a measure of similarity between left and right tibial scalp responses. For both tests, a table is presented with CCs for normal subjects. If the CC for each patient falls below the appropriate critical value from the table, then the CC is taken to be abnormal; p values are provided at specific levels in this table. Results: A new data base for the means and critical values of CCs with corresponding p values is established for normal subjects. Most of the patient data reveal significant CC nonreproducibility and lack of symmetry in tibial scalp SEP waveforms. Conclusion: The CC provides objective and quantitative waveform information that complements latency and amplitude measurements in the clinical interpretation of tibial scalp SEP waveforms.