For the prevention of the spinal cord injury in spinal surgery, intraoperative spinal cord monitoring by evoked potential is now widely accepted as a standard procedure.

Spinal evoked potential in which stimulation and/or recording is done from the spinal neural elements, has been reported to produce larger and constant signal amplitude and to be less sensitive to anesthesia than other methods. To determine the efficacy of the spinal cord monitoring, 28 cases monitored by spinal evoked potential during spine surgery were analized.

1) Changes in amplitude, less than 50% of control value were observed in four cases. Changes in latency, delayed more than 0.3 msec of control value were observed in six cases. Combined changes of both amplitude and latency were noted in three cases.

2) Temporary neurological deterioration after surgery was observed in two, in which sensory impairment showed combined changes of amplitude and latency, while motor weakness showed no significant changes.

3) The false positive change in some cases seemes to have been caused by technical or due to anesthesia, while false negative change may have resulted from changes of spinal evoked potential does not represent the changes of the motor pathway directly.

It is concluded that spinal cord monitoring by spinal evoked potential is a reliable method of detection for the deterioration of the spinal cord function during spinal surgery. However, it may be necessary to develope a method which enables the direct detection of changes in motor pathway.