Train-of-four monitoring with the twitchview monitor electctromyograph compared to the GE NMT electromyograph and manual palpation

Abstract

The purpose of this study was to compare train-of-four count and ratio measurements with the GE electromyograph to the TwitchView electromyograph, that was previously validated against mechanomography, and to palpation of train-of-four count. Electrodes for both monitors were applied to the same arm of patients undergoing an unrestricted general anesthetic. Train-of-four measurements were performed with both monitors approximately every 5 min. In a subset of patients, thumb twitch was palpated by one of the investigators. Eleven patients contributed 807 pairs of train-of-four counts or ratios. A subset of 5 patients also contributed palpated train-of-four counts. Bland–Altman analysis of the train-of-four ratio found a bias of 0.24 in the direction of a larger ratio with the GE monitor. For 72% of data pairs, the GE monitor train-of-four ratios were larger. For 59% of data pairs, the GE monitor train-of-four counts were larger (p < 0.0001). For 11% of data pairs, the GE monitor train-of-four count was 4 when the Twitchview monitor count was zero. When manual palpation of train-of-four count was compared to train-of-four count determined by the monitors, 70% of data pairs were identical between palpation and TwitchView train-of-four count, while 30% of data pairs were identical between palpation and GE train-of-four count. For 7% of data pairs, the GE monitor train-of-four count was 4 when the palpation count was 0. The GE electromyograph may overestimate the train-of-four count and ratio. The GE electromyograph frequently reported 4 twitches when none were actually present due to misinterpretation of artifacts.