Consistency of Voice Frequency and Perturbation Measures in Children

Abstract

Objective Evaluate normal pediatric voice frequency and perturbation measures with Voice Evaluation Suite (VES) and Multi Dimensional Voice Program (MDVP), determine the consistency of these measures over time, and understand which measures might be most useful for evaluating children with voice disorders. Study Design Prospective, longitudinal study of normal voices of 50 children aged 4 to 17 years. Setting Pediatric otolaryngology clinic within tertiary hospital. Subjects and Methods Two tests of sustained utterances from each child were evaluated by 2 computerized voice analysis programs for frequency and perturbation. Intraclass correlation coefficient (ICC) was used to assess the reliability between the samples. Results Children (male/female, 1.08:1) with a mean age of 8.34 years were tested on an average of 54.2 minutes apart. Each test included 4 utterances; 1 was analyzed by MDVP, and 3 grouped utterances were averaged and evaluated by VES. Fundamental frequency had excellent reliability (ICC = 0.95) in both VES and MDVP. Jitter, shimmer, and noise to harmonic ratio were poorly reliable (ICC ≤0.4) in MDVP but had good to excellent reliability (ICC 0.66-0.8) in VES. Conclusion Single, sustained utterances in children provide consistent measures of frequency. Perturbation is not reliably measured by such testing, but averaging multiple samples yields improved consistency. Evaluating acoustic measure stability in spontaneous speech and in sustained utterances cued by a tuning frequency can provide further insight on pediatric voice consistency.