The female voice in vocally demanding professions : field studies using portable voice accumulators

Abstract

Objectives: Methods to register voice use outside the voice clinic, at work, are important for assessment of work-related voice disorders. Such methods must be reliable and valid and at the same time practical and easy to handle. The overall aim of this thesis was to evaluate portable devices for long-term voice registrations and to assess voice use at work in women with vocally-demanding professions.

Materials and methods: In Study I a voice accumulator, with two different software, was evaluated regarding speaking fundamental frequency and phonation time, with reference to simultaneous studio recordings. The voice accumulator was also tested in field condition. Participants were two vocally healthy women and two vocally healthy men in both the studio and the field conditions. In Study II, the same device was tested in studio condition where one female participant simulated different voice qualities. In addition, the device was compared to simultaneous DAT recordings in field condition. Participants were three vocally healthy female preschool teachers. In Study III, the same type of voice accumulator was used to collect voice data from 12 vocally healthy female preschool teachers during two days to compare data between work and leisure. In Studies IV and V a more advanced device was used that could measure the sound pressure levels of the speaker’s voice and the environmental noise, in addition to fundamental frequency and phonation ratio. Forty women, 20 with work-related voice disorders and 20 vocally healthy controls, matched for workplace and profession, were monitored during seven days. Data comprised approximately 95 hours for each participant, categorized into work and leisure based on information from diaries. Data was analyzed with environmental noise as the experimental factor, using the device’s software program and MATLAB. The participants also rated estimated speaking time, voice symptoms and perceived disturbing noise four times a day during the week using Visual Analogue Scales. A total of 4,480 data points were collected, the response rate was 96%.

Results: Study I stressed the importance of careful placement and firm attachment of the contact microphone to the speaker’s neck for reliable detection of vocal fold vibrations. If so, the agreement between the voice accumulator’s two software and the reference studio recording was good. Field recordings indicated an activity dependent phonation time. Study II found that the voice accumulator measured fundamental frequency and phonation time reliably for different voice qualities, except for creaky voice. It also failed to register phonation at frequencies above 440 Hz as well as phonation at low sound pressure levels. Study III showed that preschool teachers used higher average fundamental frequency and phonation time during work as compared to leisure time, indicating high vocal load caused by the activities at work. Results from Study IV confirmed the finding in Study III that phonation ratio was higher during work than during leisure. When comparing results between patients and controls no significant differences were found for fundamental frequency, phonation ratio or voice level. However, significant differences were found between the patient groups; the patients with vocal nodules and their controls were exposed to significantly higher levels of environmental noise, they spoke more and louder, and used higher fundamental frequency than the patients with phonasthenia and their controls, in the work condition. Study V showed that the patients rated vocal fatigue and hoarseness as significantly higher than did the controls, as expected. The average ratings increased during the workday and remained high in the evening. A majority of the participants (32 of 40) showed significant correlations between self-rated speaking time and phonation ratio. The patients with vocal nodules and their controls rated the levels of disturbing noise as significantly higher compared to the patients with phonasthenia and their controls during workdays, which was in agreement with the instrumental measurements of the environmental noise levels.

Conclusions: The portable devices used in the present thesis were found to be useful for longtime measurements of voice use, although all had their limitations. There were no significant differences in voice use between patients and their matched controls at work, suggesting that the occupational demands and the work environment have a greater impact on vocal behaviour than individual factors. The average environmental noise reached levels that were clearly detrimental to speech communication. Therefore, a reduction of environmental noise levels seems crucial for improving the ergonomic conditions in communication-intensive and vocally-demanding workplaces. Since subjective ratings and instrumental measures had similar patterns during the week, it seems that patients and vocally healthy controls have a quite accurate perception of how much they talk and of the disturbing environmental noise. Thus, self-ratings are useful to collect, if this is done in a structured way. The difficulties encountered in long-term accumulation of data from dysphonic voices, for example at extraction and averaging of fundamental frequency of irregular vocal fold vibrations, have to be acknowledged, as does the balance between the methods’ accuracy of measurements and the need to be user-friendly.