Presentation

· Contributors · Organizations · Search Program

Vocal Loudness and Stability as Behavioral Indicators of Surgeons’ Non-Technical Skills

DescriptionSurgeons’ non-technical skills (NTS) are important for team coordination, but they are often assessed through subjective ratings. While communication has been used to assess NTS, current measures usually focus on what surgeons say, rather than how they say it. This study examined whether sentence-level vocal loudness and vocal stability can serve as behavioral indicators of surgeons’ NTS. Three surgical education fellows completed simulated surgical events designed to represent good and poor NTS. Audio recordings were transcribed and aligned at the sentence level using WhisperX. Speaker diarization was used to isolate surgeon speech. Vocal loudness was measured with root-mean-square amplitude. Each sentence was classified as low, normal, or high volume relative to the case-specific baseline. Vocal stability was measured using state entropy and state persistence probability. Compared with good NTS events, poor NTS events showed higher loudness, more frequent high-volume states, and more frequent switching between volume states. Good NTS events showed more stable and structured vocal patterns. The largest effects were observed for baseline loudness, high-volume states, state entropy, and state persistence probability. These findings suggest that sentence-level voice features may provide objective indicators of surgeons’ NTS. Future work should test these measures in larger samples and real surgical settings.

Contributors