Reliability of the perceptual evaluation of adductor spasmodic dysphonia

Although perceptual assessment by experienced voice clinicians remains the gold standard for the diagnosis and assessment of severity of adductor spasmodic dysphonia (ADSD), the interrater reliability of voice experts for this task has not been assessed. In addition, it is unknown whether telephone-recorded or -transmitted voice samples could be used for this task. The aims of this study were (1) to assess the reliability of perceptual analysis of ADSD severity by voice experts and (2) to compare the results between digitally recorded voice samples and those recorded over the telephone.

Five laryngologists randomly selected voice samples from 46 ADSD patients and rated the severity of ADSD on a 5-point rating scale. A set of digital voice recordings and a set of telephone voice recordings made from filtering the digital set via the telephone were rated, and each voice set was rated twice. Measures of intrarater and interrater reliability, as well as a measure of the probability of agreement among the raters, were calculated.

There was a high level of agreement on ADSD severity, with excellent interrater and intrarater reliability (Cronbach's alpha, .93 to .96). The probabilities of rater agreement on the digitally recorded and telephone-filtered voice samples were similar (chi2, p = .07). The ratings of digital versus telephone voice samples were highly correlated (Pearson r, 0.99; p < .001).

These results demonstrate that voice experts are reliably able to judge and agree on the severity of ADSD. Telephone-filtered voices appear to convey adequate ADSD perceptual cues for expert listeners to judge the severity of spasmodic dysphonia.