Does intraoperative nerve monitoring reduce the rate of recurrent nerve palsies during thyroid surgery?

Abstract

Two different aspects of the influence of neuromonitoring on the possible reduction of post-operative recurrent laryngeal nerve palsies require critical examination: the nerve identification and the monitoring of it's functions. Due to the additional information from the EMG signals, neuromonitoring is the best method for identifying the nerves as compared to visual identification alone. There are still no randomized studies available that compare the visual and electrophysiological recurrent laryngeal nerve detection in thyroid operations with respect to the postoperative nerve palsies. Nevertheless, comparisons with historical collectives show that a constant low nerve-palsy-rate was achieved with electrophysiological detection in comparison to visual detection. The rate of nerve identification is normally very high and amounts to 99 % in our own patients. The data obtained during the "Quality assurance of benign and malignant Goiter" study show that in hemithyreoidectomy and subtotal resection, lower nerve-palsy-rates are achieved with neuromonitoring as compared to solely visual detection. Following subtotal resection, this discrepancy becomes even statistically significant. While monitoring the nerve functions with the presently used neuromonitoring technique, it is possible to observe the EMG-signal remaining constant or decreasing in volume. Assuming that a constant neuromonitoring signal represents a normal vocal cord, our evaluation shows that there is a small percentage of false negative and positive results. Looking at the permanent recurrent nerve palsy rates, this method has a specificity of 98 %, a sensitivity of 100 %, a positive prognostic value of 10 %, and a negative prognostic value of 100 %. Although an altered neuromonitoring signal can be taken as a clear indication of eventual nerve damage, an absolutely reliable statement about the postoperative vocal cord function is presently not possible with intraoperative neuromonitoring.