A comparison of acceleromyography and mechanomyography for determination of the dose–response curve of rocuronium in children

In order to compare an acceleromyograph (TOF-Guard^TM) with a mechanomyograph (Grass FT03), the dose–response relationship of rocuronium was simultaneously determined in both arms of 15 children aged 3–11 years during anaesthesia with thiopentone, alfentanil and nitrous oxide. Three subgroups of five children received rocuronium 120, 180 or 240 μg.kg⁻¹ randomly. The effective doses to produce 50% and 95% depression of the first twitch of the train-of-four determined by acceleromyography were 206 and 337 μg.kg⁻¹, respectively, while these values determined by mechanomyography were 151 and 331 μg.kg⁻¹, respectively. The dose–response curve obtained by acceleromyography was steeper and shifted to the right compared with that obtained by mechanomyography (p < 0.0001). The difference between the effective dose producing 50% twitch depression determined by the two devices was highly significant (p < 0.0001). In 13 out of 15 children, the acceleromyograph control train-of-four ratio was significantly greater than unity. Although there was a good correlation ( r  = 0.85) between simultaneous pairs of measurements of neuromuscular block, the acceleromyograph exhibited a bias of −25% relative to the mechanomyograph with wide limits of agreement (−62 to +12%). We conclude that acceleromyographic and mechanomyographic measurements should not be used interchangeably when determining the potency of muscle relaxants.