Inducing a neuromuscular block (NMB) has advantages for the anaesthetist and the surgeon, but exposes the patient to the risk of residual curarisation.1 This can be reduced using various methods of quantitative NMB monitoring.2 The NeuroMuscular Transmission Module (M-NMT) (GE Healthcare, Hatfield, UK; formerly Datex-Ohmeda, Helsinki, Finland) uses kinemyography or electromyography (EMG). The module can record and display data throughout the period of NMB and save the data in the anaesthesia record (AS/5 monitor Gangwon-do, Korea). EMG measures muscle electrical activity using surface electrodes that can be easily attached to the hand, among other sites. The EMG responses are acknowledged to be consistent over time and less sensitive to changes in the position of the hand than acceleromyography3 and, probably, less than kinemyography.4 Given these advantages, continuous quantitative monitoring of NMB by EMG has gradually been adopted in clinical practice in our institution. Unfortunately, numerous recordings provided illogical trends of the NMB.
To illustrate the aberrant results provided by the device in some clinical cases, we report data obtained from a 95-kg patient who underwent abdominal surgery under general anaesthesia (Fig. 1). EMG monitoring was started in operating room according to the Datex-Ohmeda Neuromuscular monitoring guide5 and in accordance with good clinical practice.6 The stimulating electrodes (E152, IMMED, Bio Protech Inc., Gangwon-do, Korea) were applied to the wrist over the ulnar nerve. The recording electrode was placed on the thenar eminence over the belly of the adductor pollicis muscle and the indifferent electrode was placed on the first phalanx of the thumb. The neutral ground electrode was placed over the carpal ligament on the inner surface of the wrist. The automatic calibration sequence was completed before administering rocuronium by searching for the supramaximal stimulation level and 100% calibration of twitch height (T1). This was followed by 2-Hz train-of-four (TOF) stimulations every 20 s, with the EMG responses (TOF count, T1 and TOF ratio) recorded online.
Shortly after administering rocuronium 50 mg, four small responses reappeared on the screen in response to the TOF stimulations. The calculated TOF ratio was more than 0.9. Surprisingly, this figure fell gradually as the T1 height increased. The EMG responses did not disappear after further boluses of rocuronium 10 mg, and the TOF ratio remained high until surgery ended. Sugammadex 200 mg enabled rapid recovery of four responses with T1 more than 100% and without fade (TOF ratio > 0.9) and made it possible to extubate the patient in the absence of residual block.
The main problem was the unexpected recording of the four small responses that persisted during deep and moderate blockade (Fig. 1b). Only two actual muscular EMG responses were recorded within the four identical artefacts. However, the device interpreted them as four real twitches (T1: 12% and TOF count: 4; Fig. 1c). The TOF ratio was inappropriately calculated (83%, GE Healthcare) and increased after further administration of rocuronium.
To determine how frequently this problem arises in our clinical practice, we carried out a chart review to objectively assess the quality of the EMG recordings provided by the Datex-Ohmeda M-NMT module. Following Institutional Ethics Committee approval (CHU Mont-Godinne om 050, 69/2011), 100 EMG recordings of non-depolarising NMB randomly selected over a period of 1 year were analysed retrospectively. The patients’ mean (SD) age was 51 (19) years and BMI was 26 (5) kg m2 and they underwent various types of surgery. Rocuronium 0.48 (0.11), atracurium 0.45 (0.10) and cisatracurium 0.13 (0.05) mg kg were administered to 68, 30 and two patients, respectively. Neostigmine was administered in 45% of cases.
The initial EMG was properly calibrated in 72 cases. Taking into account whether or not the initial calibration was carried out, the quality of various aspects of the EMG recordings was assessed: the capacity to determine the TOF count gradually (from 1 to 4) – 82 and 54% of cases, respectively; whether the early trend of TOF ratio was considered consistent (gradually increasing from an initial level <50%, and inter-measurement variability <15%) – 34 and 25% of cases, respectively; and, conversely, if the TOF ratio's final assessment was consistent – in 91 and 71% of cases, respectively.
We have shown that when using the M-NMT Datex-Ohmeda module, problems may exist regarding the interpretation of the EMG recordings to the point where this limited their usefulness in managing the NMB. In particular, the detection of four small, almost identical EMG responses shortly after the induction of the block resulted in an incorrect calculation of a TOF count and a high TOF ratio. This problem frequently led to the false assumption that the patient was not completely paralysed (up to 75% of cases in the absence initial calibration). The presence of these small responses without fade during deep block was previously attributed to direct muscular stimulation. However, recent studies have shown that the current used in NMB monitoring (pulse duration 0.2 ms, intensity 10–60 mA) was usually inadequate to induce genuine muscular stimulations.7 Thus, other reasons specific to the method of EMG recording used by the Datex-Ohmeda M-NMT module probably contributed to the artefactual detection of these small responses. Further prospective investigation should be undertaken to determine the cause.
During NMB recovery, as ‘real’ twitches appear and then increase, they replace the artefacts and the calculated TOF ratio decreases progressively (as the real twitches exhibit fade), before increasing again until as fade disappears towards the end of recovery (Fig. 1a). The quality of the TOF ratio measurements increased with the intensity of the muscular responses recorded, reducing the impact of artefacts at the end of the case. However, in 29% of cases when initial calibration had not been performed, it was not possible to rule out residual curarisation before the patient awoke. We have shown that the recordings were of a better quality when calibration was undertaken prior to administering rocuronium. Even if artefacts remain, it seems useful to calibrate the initial twitches to limit the magnitude of electrical interference during the recording of the muscular responses.
In conclusion, artefacts alter EMG recordings provided by the Datex-Ohmeda M-NMT module, particularly during deep and moderate block. Additional technical improvements are needed to guarantee clinicians correct management of NMB in all patients.
This work was supported by departmental funds.
A.d’.H. is a non-remunerated consultant and board member of Medical Devices Bio Engineering-sprl, Waterloo, Belgium.
1. Murphy GS, Brull SJ. Residual neuromuscular block: lessons unlearned. Part 1. Definitions, incidence, and adverse physiologic effects of residual neuromuscular block. Anesth Analg
2. Brull SJ, Murphy GS. Residual neuromuscular block: lessons unlearned. Part II. Methods to reduce the risk of residual weakness. Anesth Analg
3. Hänzi P, Leibundgur D, Wessendorf R, et al. Clinical validation of electromyography and acceleromyography as sensors for muscle relaxation. Eur J Anaesthesiol
4. Dahaba AA, Von Klobucar F, Rehak PH, List WF. The neuromuscular transmission module versus the relaxometer mechanomyograph for neuromuscular block monitoring. Anesth Analg
5. Brull SJ, Paloheimo M. A practical guide to monitoring neuromuscular function. Datex-Ohmeda Division Instrumentarium Corp. Helsinki, Finland; 2002.
6. Fuchs-Buder T, Claudius C, Skovgaard LT, et al. Good clinical research practice in pharmacodynamic studies of neuromuscular blocking agents II: the Stockholm revision. Acta Anaesthesiol Scand
7. Kopman AF. Can conventional peripheral nerve stimulators induce direct muscle depolarization? Anesth Analg