Validation of a tablet-based application for hearing self-screening in an adult population.

OBJECTIVE: This study evaluated the diagnostic performances of a tablet-based hearing screening test by assisted-test and self-test modes. DESIGN/METHOD: Measurements were performed with the SoTone tests in normal hearing and hearing-impaired adult participants using an Android tablet and calibrated Bluetooth headphones. The duration of assisted- and self-test modes were compared. Receiver operating characteristic (ROC) curves were plotted after calculations of sensitivity and specificity at 20, 30, and 35 dB HL cut-off values. STUDY SAMPLE: 217 participants performed the tests. The effect of test mode (assisted versus self) was compared in a sample of 103 participants. RESULTS: Self-test duration (89 s) was significantly longer than the assisted-test duration (75 s) (p = 0.003, Wilcoxon test). For the 20, 30, and 35 dB HL cut-off values, sensitivity was between 92% and 96%, and specificity was between 79 and 90%. Concordance of results between assisted-test and self-test modes was excellent (Cohen's kappa = 0.81, p < 0.001). CONCLUSIONS: The SoTone hearing screening test is accurate for identifying the presence of a suspected hearing loss at 20 dB HL or more in adults. It can be used either in assisted-test or self-test modes.

Speech perception, real-ear measurements and self-perceived hearing impairment after remote and face-to-face programming of hearing aids: A randomized single-blind agreement study.

INTRODUCTION: Current literature does not provide strong evidence that remote programming of hearing aids is effective, despite its increasing use by audiologists. We tested speech perception outcomes, real-ear insertion gain, and changes in self-perceived hearing impairment after face-to-face and remote programming of hearing aids in a randomized multicentre, single-blind crossover study. METHODS: Adult experienced hearing aid users were enrolled during routine follow-up visits to audiology clinics. Hearing aids were programmed both face to face and remotely, then participants randomly received either the face-to-face or remote settings in a blinded manner and were evaluated 5 weeks later. Participants then received the other settings and were evaluated 5 weeks later. RESULTS: Data from 52 out of 60 participants were analysed. We found excellent concordance in performance of hearing aids programmed face to face and remotely for speech understanding in quiet (phonetically balanced kindergarten test - intraclass correlation coefficient of 0.92 (95% confidence interval: 0.87-0.95)), and good concordance in performance for speech understanding in noise (phonetically balanced kindergarten +5 dB signal-to-noise ratio - intraclass correlation coefficient of 0.71 (95% confidence interval: 0.55-0.82)). Face-to-face and remote programming took 10 minutes (±2.9) and 10 minutes (±2.8), respectively. Real-ear insertion gains were highly correlated for input sound at 50, 65 and 80 dB sound pressure levels. The programming type did not affect the abbreviated profile of hearing aid questionnaire scores. CONCLUSIONS: In experienced hearing aid users, face-to-face and remote programming of hearing aids give similar results in terms of speech perception, with no increase in the time spent on patients' care and no difference in self-reported hearing benefit. CLINICALTRIALS.GOV IDENTIFIER: NCT02589561.