The TOF-Guard neuromuscular monitor uses an accelerometer to measure t
he response to nerve stimulation. In this study, we have compared it t
o a standard mechanomyographic monitor, the Myograph 2000, for neuromu
scular monitoring in 28 subjects. A train-of-four mode of stimulation
was used in both cases. The times taken for onset of block, and for th
e recovery of T1 (the first response in the train of four) to 25% of c
ontrol, the time from recovery of T1 from 25-75% and for the recovery
of the train of four ratio to 0.7 were compared with the two monitors.
There was a good correlation between the two devices for both onset a
nd recovery times. However, differences were highlighted when the data
were analysed by the method of Bland and Altman. The 95% limits of ag
reement for the T1 recovery to 25%, as measured by the TOF-Guard, rang
ed from 5 min less to 8 min more than when measured by the Myograph 20
00. For recovery of the train of four ratio to 0.7, the limits of agre
ement were approximately 6 min in either direction. The 95% limits for
the TOF-Guard measured train of four ratio were from 0.47 to 0.99, at
the Myograph reading of 0.7. We recommend that information from the T
OF-Guard and the Myograph 2000 should not be used interchangeably. How
ever, the TOF-Guard is likely to improve considerably on tactile evalu
ation of the responses to stimulation.