Latency shift in compound muscle action potentials during electroneurography in facial palsy

Objective Electroneurography (ENoG) reliably predicts the prognosis of facial palsy. However, the results of ENoG are dependent on the location, where the wave is detected, as a compound muscle action potential (CMAP) arising from the facial muscles. To minimize errors in prognostic prediction, we analysed the latencies of facial CMAPs. Materials and methods Fifty-seven patients with unilateral peripheral facial palsy and 24 healthy volunteers were enrolled. Amplitudes, negative peak latencies (NPL), and rise latencies (RL) of CMAPs were measured on the paralysed and healthy sides in patients and in healthy volunteers. The relationships of these latencies with ENoG values and the lowest House-Brackmann (H-B) scores were also analysed. Results The amplitude of CMAP on the paralysed side was smaller, and NPL and RL were longer, than those on the healthy side in patients and healthy volunteers (p < 0.01). In patients, there was no difference in NPL between the ENoG < 40% group and the ENoG >= 40% group. Conversely, there was a significant difference in RL between the ENoG < 40% group and ENoG >= 40% group (p = 0.03). No relationships were observed between NPL or RL and the lowest H-B score. Conclusions NPL and RL of CMAP on the paralysed side were equivalent or longer than those on the healthy side. During ENoG for facial palsy, CMAP should be measured on the healthy side first, and then detected (and the amplitude measured) on the paralysed side with reference to CMAP latency on the healthy side, to reduce errors in detecting facial CMAPs.