Effects of Personal Protective Equipment on Speech Acoustics

Abstract

Objectives: The transmission of severe acute respiratory syndrome coronavirus-2 occurs primarily through droplets, which highlights the importance of protecting the oral, nasal, and conjunctival mucosas using personal protective equipment (PPE). The use of PPE can lead to communication difficulties between healthcare workers and patients. This study aimed to investigate changes in the acoustic parameters of speech sounds when different types of PPE are used. Methods: A cross-sectional study was conducted, enrolling 18 healthy male and female participants. They were instructed to produce a sustained [a:] vowel for at least 3 s to estimate voice quality. In addition, all Turkish vowels were produced for a minimum of 200 ms. Finally, three Turkish fricative consonants ([f], [s], and [f]) were produced in a consonant/vowel/consonant format with different vowel contexts within a carrier sentence. Recordings were repeated under the following conditions: no PPE, surgical mask, N99 mask, face shield, surgical mask + face shield, and N99 mask + face shield. All recordings were subjected to analysis. Results: Frequency perturbation parameters did not show significant differences. However, in males, all vowels except [u] in the first formant (F1), except [(sic)] and [u] in the second formant (F2), except [epsilon] and [(sic)] in the third formant (F3), and only [i] in the fourth formant (F4) were significant. In females, all vowels except [i] in F1, except [u] in F2, all vowels in F3, and except [u] and [.] in F4 were significant. Spectral moment values exhibited significance in both groups. Conclusion: The use of different types of PPE resulted in variations in speech acoustic features. These findings may be attributed to the filtering effects of PPE on specific frequencies and the potential chamber effect in front of the face. Understanding the impact of PPE on speech acoustics contributes to addressing communication challenges in healthcare settings.