Acoustic Psychometric Severity Index of Dysphonia (APSID): Development and Clinical Application

Acoustic Psychometric Severity Index of Dysphonia (APSID): Development and Clinical Application:



Publication date: Available online 4 December 2019

Source: Journal of Voice

Author(s): Seung Jin Lee, Hong-Shik Choi, HyangHee Kim

Summary

Objectives/Hypothesis

The purpose of this study was to explore the predictability of the Korean version of the Voice Activity and Participation Profile (K-VAPP) scores and acoustic measures for perceived severity of voice disorders. Furthermore, usefulness of the index derived from the multivariate formula as a screening tool for voice disorders was explored and compared with that of the Cepstral Spectral Index of Dysphonia (CSID).

Methods

A total of 150 patients with various voice disorders (42 males and 108 females, age = 45.15 ± 15.15 years) and matched 50 normal controls (14 males and 36 females, age = 44.78 ± 13.70 years), participated in the study. Correlates of perceived overall severity were explored by a stepwise regression analysis with regards to demographic information (age, gender, and professional voice use), jitter, shimmer, and noise-to-harmonic ratio of vowel samples, cepstral peak prominence (CPP), L/H spectral ratio, and their standard deviations of sentence and vowel samples, and psychometric measures of the K-VAPP. Using the ROC curve analysis, the area under curve (AUC) of the novel index derived from the regression equation and the CSID were compared and the cut-off score of the index was obtained.

Results

A significant portion (adjusted R square = 79.0%) of the overall severity perceived by clinicians was predicted by five parameters: CPP of the sentence and vowel production, the self-perceived severity subscale score of the K-VAPP, σCPP of the vowel production, and gender. The AUC of the novel index was 0.935 (95% confidence interval 0.891–0.965) and significantly higher than that of the CSID of vowel samples (AUC = 0.832). The cutoff score with the highest Youden J Index (0.7612) was >22.257617 (sensitivity = 82.12%, specificity = 94.00%).

Conclusion

The current data showed the possibility that a novel index reflecting both acoustic and psychometric correlates of auditory-perceptual estimates for dysphonia severity could be suggested and possibly named as the “Acoustic Psychometric Severity Index of Dysphonia (APSID).” Further studies pertaining to the follow-up of various voice disorders are needed to extend its clinical usefulness.