Abstract:
In France 58% of persons with hearing loss still do not wear hearing aids. Pure-tone
audiometry is the traditional gold standard in assessment and screening of hearing
impairment, but it requires the use of calibrated devices and soundproof booth. The
antiphasic digits-in-noise (DIN) test does not require calibrated material and can run on
a standard headset or earbuds connected to a smartphone or a computer. The DIN test
is highly correlated with pure tone audiometry and has already shown to be effective
in hearing loss screening in its English version promoted by the WHO. The aim of the
present study was to develop and validate a French version of the antiphasic DIN test
for implementation on a national screening test offered as a smartphone app. The audio
files recorded from a French native female speaker were selected and normalized in
intensity according to their recognition probability. The French DIN test application was
then tested on normal hearing- and hearing-impaired subjects. Based on the strong
correlation between pure tone audiometry (PTA) and DIN SRT, we calculated ROC curves
and Z-score. For PTA > 20 dB HL, a SNR cutoff of 12.9 dB corresponds to a sensitivity
and specificity of 0.96 and 0.93, respectively. To detect moderate and more severe
hearing loss (PTA > 40 dB HL), the SNR cutoff was −10.9 dB, corresponding to a
sensitivity and specificity of 0.99 and 0.83, respectively. The Z-score was calculated to
define statistical criteria of normality for speech-in-noise evaluation. While a score of 0
roughly corresponds to the normality (DIN SRT = −15.4 dB SNR), a subject with DIN
SRT > −12.2 (Z-score > 2) is ranked in the hearing loss population. Next, the French
antiphasic DIN test was implemented in the Höra iOS and Android apps. In total, 19,545
Höra tests were completed and analyzed. Three quarters of them were classified as
normal (74 %) and one quarter presented mild (9%) or more severe loss (17%). Together, results argue for the use of the French version of antiphasic DIN test in the general
population to improve the screening of hearing-impaired individuals.
