The accuracy of a smartphone-based hearing assessment compared to conventional audiometry in workers exposed to noise according to sociodemographic, occupational and pre-assessment conditions.
Claire Márcia Santana Lima, Luma Cordeiro Rodrigues Fernandes, Ana Paula Corona
The hearTest smartphone hearing test was accurate for most noise-exposed workers, but fell below the 70% accuracy threshold for workers aged 40–49, white workers, and those with elementary education or less.
Key findings
1The hearTest met the 70% accuracy threshold for identifying any hearing loss in most worker subgroups, but fell below it for workers aged 40–49 (J = 51.6%), white workers (J = 62.5%), those with elementary education or less (J = 56.4%), and those earning more than 5 minimum wages (J = 66.7%).
2For identifying noise-induced hearing loss specifically, the hearTest's accuracy fell below the 70% threshold for workers aged 40–49 (J = 64.5%), those with elementary education or less (J = 61.2%), those earning below 2 minimum wages (J = 67.3%), and those who reported irregular sleep the night before (J = 66.2%).
3The study re-analysed data from 232 noise-exposed workers (94.4% male, 94.8% non-white, 80.2% in industrial roles) who completed both the hearTest and conventional audiometry on the same day at an occupational health clinic.
Still to come
How it was doneWhat they foundWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The sample was 94.4% male, 94.8% non-white, and 80.2% in industrial roles, so findings may not generalise to women, white workers, or office-based populations. Only 13 women and 7 administrative workers were included, making subgroup accuracy estimates for these groups unstable or uncalculable. This is a secondary analysis of previously collected data; the researchers did not design the original study or control its procedures. The hearTest was always administered before conventional audiometry, with no counterbalancing, so order effects (fatigue, learning) cannot be separated from true accuracy differences. The hearTest was conducted in a soundproof room without specific acoustic treatment, and the authors acknowledge that residual environmental noise may have affected low- and mid-frequency thresholds. Variables were analysed one at a time rather than in a multivariable model, so the independent contribution of each factor (e.g., education versus income versus occupation) cannot be determined.
Declared interests
Funded by the Brazilian National Council of Technological and Scientific Development (CNPq, Grant 429093/2016-0) and the Federal University of Bahia (Grant 004/2016). No industry sponsorship, device-manufacturer involvement, or author conflicts of interest were declared.
The easy way to misread this
Do not read the subgroup accuracy differences (for example, the lower Youden Index for white workers or 40–49-year-olds) as proof that the hearTest is unreliable for those groups. Several subgroups contained very few participants (12 white workers, 7 administrative workers, 13 women), and some accuracy measures could not be calculated at all because no one in the subgroup had hearing loss on the reference test. The differences may reflect small numbers and confounding by occupation rather than a true device limitation.
Summarised by AI from the full paper, without a clinician reviewing it. Check it against the source before it changes what you do. Read it on PubMed →