Speech-Based AI for Stress : Uncovering Cigna’s Toolkit

Reading Time: < 1 minute

Cigna’s stress toolkit features the Cigna StressWaves Test (CSWT), an AI tool assessing psychological stress through speech analysis. This study independently validates the CSWT, revealing low repeatability and convergent validity. The public release of the CSWT without ample validation data raises concerns about deploying digital health tools prematurely.

Psychological stress links to global health issues like cardiovascular disease, hypertension, and depression1,2. Traditionally, stress measured with the Perceived Stress Scale (PSS), a reliable and valid instrument3,4,5,6. Recently, AI-based tools, including the CSWT, gained interest for stress assessment11,12. Despite its global use, the CSWT lacks published validation data. This paper provides independent validation for the CSWT.

Speech-based AI for stress lacks validation compared to traditional tools. Established scales like the PSS show high reliability and validity3,4,5,6.

Despite lacking validation, the CSWT claims “clinical-grade” performance16 and serves as a “stress diagnostic tool”17. This paper assesses these claims by examining the CSWT’s test–retest reliability and validity against the PSS.

Source Nature

Author: Neurologica