Recording of somatosensory evoked potential (SSEP) responses is achieved through the stimulation of afferent peripheral nerves and recording the responses elicited by this stimulation, mitigated proximally via the ascending sensory fibers in the dorsal column to be processed in the primary sensory cortex.

Since the dorsal sensory tracts are to be engaged, theoretically any (somatic) stimulation would be effective in provoking the desired ascending volley. One might reach under the drape and continually tap the ankle posterior to the medial malleolus and be able to record an SSEP response. However, practicality demands a stimulation technique that is simple, effective, and efficient. This is achieved through the use of surface electrodes to deliver an electrical current of a given frequency and intensity to the skin area superficial to large sensory nerves. Common sites for this stimulation are the posterior tibial nerve (PTN) in the ankle, and the ulnar or median nerve in the wrist, with sites at the common peroneal notch and ulnar groove as less frequent alternatives.

Recording sites should include something peripheral as a control to indicate that the stimulation has effectively “gotten in,” such as the popliteal fossa for PTN stimulation, and the supraclavicular notch for recording an Erb’s point response from median or ulnar nerve stimulation. Parameters for stimulating and recording settings are well established and are presented in Table 68.1.

The responses generated from somatosensory stimulation are exceedingly small, dwarfed by other physiologically generated electrical signals such as the cardiac node pulse and even the passive EEG activity of the brain. These other organic generators act as a source of electrical “noise” in a recorded SSEP just as would artificial 60 Hz contamination. However, because SSEP responses occur at the same time relative to each stimulation , a mathematic technique using signal averaging may be employed to resolve the response signal out from the background noise. Since the noise generated by other sources tends to be of a random polarity for a given time period, over the course of many mathematical averages the time-locked response will additively increase while the random activity will cancel itself out. Ultimately, after a number of repetitions, this allows for a signal-to-noise ratio that is adequate to appreciate the SSEP response.

Primary interpretation of the SSEP involves measuring the amplitude and latency of the responses. Because these responses are going to be used as their own control, rather than compared to a theoretical norm based on absolute values, it is essential that preincision, or premanipulation baselines are established. Discovering a significantly abnormal latency or amplitude value after the fact, without an established baseline generated prior to any surgical risk being incurred to compare to, there is no value to the surgeon or to the patient. Indeed, significance is only established as a delta between the baseline values and the ones recorded during the surgical procedure.

The well-known criterion for determining significance in latency shift is an increase of 10% or more from baseline. For amplitude, significance is reached after a 50% or greater amplitude reduction from baseline. It is important to remember that anesthetic drugs commonly used during surgery can alter SSEP responses by both increasing latency and decreasing amplitude.

As stated previously, a known limitation of SSEP is the fact that it cannot directly measure the integrity of the descending corticospinal motor pathways. Rather, in an SSEP-only paradigm, indirect inference must be made concerning the integrity of the anterior and lateral motor pathways; such inference is fraught with potential danger. Indeed, reports of postoperative motor dysfunction in the context of preserved intraoperative SSEP are not uncommon. Dinner et al. report a 2% incidence of false-negative SSEP detection for new-onset motor deficit [4]. Because general consensus reveals that a motor deficit is regarded as a much more severe sequela than a sensory loss, Motor Evoked Potentials (MEP) were therefore developed as a way of more comprehensively monitoring the spinal cord by including the vital motor pathways within the neurophysiologists scrutiny.

Typically, the stimulating electrodes are placed over the C3 and C4 locations, based on the standardized international 10–20 measuring system. This scalp location puts them in close proximity to the precentral motor gyrus, beneath the bone of the skull. A more medial placement, designated as C1 and C2, is located halfway between C3/C4 and Cz. Using these placement locations may help to improve responses, especially from the lower extremities. Recording electrodes are typically placed in muscle, thus allowing for the recording of a compound muscle action potential (CMAP) as the derived response. In determining which muscles to record from, consideration should be given to the expected spinal levels at risk. Where possible, a TcMEP response should be recorded from a muscle whose innervation is not directly involved with the operative level, in order to act as a control relative to responses at and distal to the at-risk levels.

Delivering 200 V or more to the surface of the head does more than produce a descending corticospinal depolarization. Superficial muscles in the scalp, face, and neck are also activated. This can lead to moderate to strong contraction of the mastication muscles, essentially creating a bite. Therefore, every care should be taken to avoid the adverse side effect of tongue or lip laceration due to this motor activation. Soft bite blocks, such as rolled 4 × 4 gauze pads should be placed bilaterally between the molars so that no travel is possible between the maxilla and mandible. Even in edentulate patients, bite blocks are recommended so as to prevent bruising of the lips or gums on the endotracheal tube .

Criteria for reporting a significant change in TcMEP responses are not yet as readily agreed upon as for those of SSEP. Currently, two major proposed paradigms vie for consensus: the so-called presence–absence (PA) model and the stimulus threshold (ST) model .