Relative effectiveness of electrically- vs mechanically-elicited EMGs in detecting pedicle wall perforation and surgically-induced nerve root damage.

Electrical stimulation of a pedicle hole and screw with recording EMGs from the lower extremities has been used as an indicator in detecting perforations of the pedicle. Mechanically-elicited EMGs are reported to be sensitive to mechanical irritation of nerve roots. This study analyzed the sensitivity of the data elicited by two EMG monitoring methods in the presence of a neurologic deficit caused by a malpositioned screw to determine the relative effectiveness of electrically- vs mechanically-elicited EMGs in detecting pedicle wall perforations and nerve root damage in patients undergoing spinal surgery utilizing transpedicular instrumentation. One hundred and four surgeries were monitored using the two EMG methods. Six hundred and fifty-four pedicle holes were prepared and 650 placed pedicle screws were electrically tested. Mechanically-elicited EMGs were monitored from a total of 618 muscles. Electrically-elicited EMGs showed a 62% true-positive rate and a 0.2% false-negative rate in detecting pedicle wall perforations. None of the patients who initially demonstrated abnormal electrically-elicited EMGs demonstrated any post-operative neurologic problems due to an incorrect screw placement. Only one patient who had abnormal mechanically-elicited EMGs during the procedures related to instrumentation developed new L4 radiculopathy immediately post-operatively which was consistent with the level of mechanically-elicited EMGs. Mechanically-elicited EMGs showed a 100% true-positive rate for nerve root irritation and a 3.5% false-negative rate in detecting pedicle wall perforations by malpositioned screw. In conclusion, although mechanically-elicited EMGs were an insensitive technique in detecting a perforation of the pedicle, mechanically-elicited EMGs were more beneficial than electrically-elicited EMGs in detecting the risk of nerve root irritation.