Inter-rater reliability of seven neurolaryngologists in laryngeal EMG signal interpretation.

Laryngeal electromyography (LEMG) has been considered as gold standard in diagnostics of vocal fold movement impairment, but is still not commonly implemented in clinical routine. Since the signal interpretation of LEMG signals (LEMGs) is often a subjective and semi-quantitative matter, the goal of this study was to evaluate the inter-rater reliability of neurolaryngologists on LEMGs of volitional muscle activity. For this study, 52 representative LEMGs of 371 LEMG datasets were selected from a multicenter registry for a blinded evaluation by 7 experienced members of the neurolaryngology working group of the European Laryngological Society (ELS). For the measurement of the observer agreement between two raters, Cohen's Kappa statistic was calculated. For the interpretation of agreements of diagnoses among the seven examiners, we used the Fleiss' Kappa statistic. When focusing on the categories "no activity", "single fiber pattern", and "strongly decreased recruitment pattern", the inter-rater agreement varied from Cohen's Kappa values between 0.48 and 0.84, indicating moderate to near-perfect agreement between the rater pairs. Calculating with Fleiss' Kappa, a value of 0.61 showed good agreement among the seven raters. For the rating categories, the Fleiss' Kappa value ranged from 0.52 to 0.74, which also showed a good agreement. A good inter-rater agreement between the participating neurolaryngologists was achieved in the interpretation of LEMGs. More instructional courses should be offered to broadly implement LEMG as a reliable diagnostic tool in evaluating vocal fold movement disorders in clinical routine and to develop future algorithms for therapy and computer-assisted examination.