Evoked potentials: anesthetic effects


The purpose of intraoperative evoked potentials is to monitor neural pathways in hopes of avoiding iatrogenic injury to the nervous system. Sensory evoked potentials (SEPs) evaluate the integrity of ascending sensory tracts while motor evoked potentials (MEPs) deal with the functionality of descending motor pathways.


Three SEP modalities are used clinically: somatosensory (SSEP), auditory (BAEP), and visual (VEP). The amplitude of these potentials is much smaller than that of the background EEG and must be extracted from the surrounding “noise” using signal averaging. Compromise of a neurologic pathway is evident with an increase in the latency and/or a decrease in the amplitude of EP waveforms. All anesthetics influence EP waveforms to some degree.

EPs of cortical origin (i.e. the cortical portion of SSEPs and VEPs) are considered more prone to modification by anesthetics than brainstem potentials (i.e. BAEPs and subcortical portions of SSEPs). It is imperative to maintain constant anesthetic drug levels during recording of EPs. Bolusing IV infusions or abrupt changes in MAC values of inhaled agents can be detrimental, particularly during critical portions of a procedure. When recording cortical EPs (SSEPs and VEPs) one should generally incorporate intravenous techniques as high concentrations of volatiles can completely eliminate cortical EPs. Nevertheless, all volatile anesthetics increase cortical latency and decrease cortical amplitude. However, desflurane and sevoflurane appear to be more forgiving with SSEPs than agents such as enflurane and isoflurane. For instance, desflurane at up to 1 MAC without nitrous oxide is compatible with cortical median nerve SSEP monitoring during scoliosis surgery. Use of nitrous oxide should probably be avoided during EP monitoring as it exhibits profound depressant effects on SSEPs and VEPs when combined with a volatile agent.

SSEPs: Pathway = Dorsal columns (UE)/lateral funiculus (LE). Electrical stimulation of mixed nerve, cortical short latency waveforms measured. Decreased amplitude of 50% + 10% prolongation in latency are significant. Signal averaged. Volatiles = dose dependent decrease in amplitude, increase in latency (< 1.0 MAC is ideal, not required)

Brainstem potentials are quite resilient with regard to anesthetics and are considered compatible with most anesthetic regimens.

Regarding IV agents, studies on thiopental have shown that even with doses far in excess of what is required to produce an isoelectric EEG, SSEP and BAEP waveforms are preserved even if there is a predictable increase in latency and decrease in amplitude. Propofol increases latency and decreases amplitude of cortical EPs, but is considered an integral component of balanced intravenous neurosurgical anesthesia. Opioids produce minimal changes in SEP waveforms, even in high doses. Because of this, they are recommended for use as infusions during EP monitoring. Both clonidine and dexmedetomidine decrease anesthetic requirements and have minimal effects on cortical EPs and are considered safe to use during monitoring. For practical purposes, ALL intravenous agents have negligible effect on cortical SSEP’s, except for etomidate and ketamine, which can actually increase amplitude. Propofol, in concentrations sufficient to induce burst suppression, will abolish later cortical responses, but not the primary cortical responses (N20/P22 for median nerve, P40/N45 for tibial nerve), which are used for monitoring.


Transcranial electrical stimulation of the motor strip is a reliable method of producing intraoperative motor potentials that can be recorded from the spinal cord, peripheral nerve, or muscle. MEPs are exquisitely sensitive to anesthetics, especially inhalational agents. Volatiles, therefore, should be avoided during recording of myogenic MEPs. Benzodiazepines, barbiturates, and propofol all depress MEPs, however, adequate recordings can be obtained during propofol anesthesia by controlling serum levels and increasing stimuli rates. Muscle relaxants can affect the recorded EMG response by depressing myoneural transmission. However, adequate MEP recordings can be achieved as long as one or two twitches on TOF can be maintained.

Keyword history

  • 55%/2008
  • 83%/2010