Performance of a new auxiliary device based on wrist brace to improve accuracy and feasibility in neuromuscular monitoring with acceleromyography in prone-positioned patients undergoing lumbar spine surgery: a prospective randomized clinical trial.

Accuracy of acceleromyography (AMG) is not be comparable to that of mechanomyography or electromyography (EMG). In particular, the prone position may reduce the accuracy and feasibility of AMG. We developed a new device based on wrist brace to allow free thumb movement and fix the other parts of the hand and wrist. We aimed to test whether the brace applied to the AMG would increase precision of AMG and agreement with the EMG in the prone position. Fifty-seven patients, undergoing lumbar surgery under general anesthesia, were randomly assigned to groups to which AMG was applied with or without (no) brace (29 in group B vs. 28 in group NB). EMG was performed in the contralateral arm. Repeatability coefficients of the first twitch height (T1) and train-of-four (TOF) ratio were assessed from nine consecutive measurements during spontaneous recovery from rocuronium-induced neuromuscular block and the AMGs of the two groups were compared in prone position. The agreement between AMG and EMG in each group was assessed using the Bland-Altman method. In group B, the repeatability coefficient of T1 was significantly lower during the recovery to T1 of 25% and TOF ratio of 0.9 (P = 0.017 and 0.033, respectively), indicating higher precision. The mean differences of bias (95% limits of agreement) between AMG and EMG in TOF ratio of 0.9 were 6.839 (- 26.54 to 40.22) in group NB and 3.922 (- 21.83 to 29.67) in group B. The wide limits of agreement in group NB was slightly narrowed in group B but without significance. Trial registration: registered on the UMIN Clinical Trials Registry in August 2020 (UMIN000041310).