Surface electrodes are not sufficient to detect neurotonic discharges: observations in a porcine model and clinical review of deltoid electromyographic monitoring using multiple electrodes.

OBJECTIVE: In order to define the preferred electromyographic monitoring method during spine surgery, (1) a porcine model of neurotonic generation after lumbar root compression was developed and (2) intraoperative use of deltoid muscle intramuscular needle, subdermal needle, and surface electrodes was retrospectively reviewed.
METHODS: In pigs, an array of intramuscular needle, subdermal needle, and surface electrode derivations was differentially amplified at identical gain and filter settings. Nerve root compression generated neurotonic discharges whose amplitudes were compared at each derivation. Clinically, 25 deltoid muscles in 13 patients were simultaneously monitored (during cervical spine surgery at the C4-C5 level) with surface, subdermal needle, and intramuscular needle electrode pairs, differentially amplified at identical gain and filter settings. Non-repeating neurotonic discharges were assigned, by amplitude and morphology, to best derivation (intramuscular, subdermal, surface or combination); coincident amplitudes were measured at the maximum deflection among the three derivations. Actual voltage detected between clinical methods was analyzed with Friedman's test and any detection versus none by general estimating equations(GEE) using SAS. The advantage of two needles over one in detection of any voltage was assessed using McNemar's test.
RESULTS: Compressed porcine lumbar roots generated neurotonics which were identifiable at intramuscular sites only. Clinically, 31 neurotonics were identified: 20/31 at intramuscular, 5/31 at subdermal, and 6/31 equally well at intramuscular and subdermal derivations. Intramuscular detected neurotonics better than subdermal derivations (z = 2.9, P < .004). No voltage was recorded at the surface in 16/31 neurotonics. For detection of any voltage, intramuscular was better than subdermal (z = -1.5, P = .04) or surface electrodes (z = -2.7, P < .001).
CONCLUSIONS: Electromyographic moni- toring of spine surgery should not be done by surface electrodes. Because sensitive neurotonic detection requires near field recording, intramuscular electrodes are preferred. Monitoring of a myotome at particularly increased risk may suggest multiple intramuscular electrodes.