Using Transcutaneous Laryngeal Ultrasonography (TLUSG) to Assess Post-thyroidectomy Patients' Vocal Cords: Which Maneuver Best Optimizes Visualization and Assessment Accuracy?

Abstract

To assess vocal cord (VC) movement with transcutaneous laryngeal ultrasound (TLUSG), three maneuvers, namely passive (quiet respiration), active (phonation), and Valsalva maneuvers have been described. It remains unclear which maneuver or using more maneuvers provides better visualization and assessment accuracy. We prospectively evaluated 342 post-thyroidectomy patients from two centers. They underwent TLUSG with direct laryngoscopic (DL) validation afterwards. During TLUSG, patients were instructed to perform all three maneuvers (passive, active, and Valsalva). VC visualization rate and accuracy between three maneuvers were compared. Visualization rate tended to be higher in Valsalva maneuver than that in other two maneuvers (92.1% vs. passive: 91.5%; active: 89.8%). While 19 patients had post-operative VC palsy, passive maneuver had lower test specificity than active (94.3 vs. 97.6%, p = 0.01) and Valsalva maneuvers (94.3 vs. 97.4%, p = 0.02). In assessable VCs, passive maneuver has a higher ability to differentiate between mobile VCs and VC palsy (Area under ROC curve--passive: 0.942, active: 0.863, Valsalva: 0.893). TLUSG with more maneuvers did not improve sensitivity or specificity. On applying TLUSG as a screening tool (i.e., only selected patient with "unassessable" VCs or VCP on TLUSG for DL), Valsalva maneuver (85.96%) saved more patients from DL than passive (81.87%) or active (84.81%) maneuver. Passive maneuver has a higher ability to differentiate VC palsy from normal. Using TLUSG as a screening tool, Valsalva was the preferred maneuver as it was more specific, had high visualization rate, and saved more patients from DL.