Authors: Yan Jiang¹, Paula Folkeard¹, Arman Hassanpour¹, Vijay Parsa¹, Ewan Macpherson¹, Susan Scollie¹, Prudence Allen¹
¹Western University
Background: Not every user experiences the same speech-in-noise perception benefits from hearing aids, even if they have similar audiograms. What best predicts individual hearing aid (HA) benefits? In this study, we examined potential predictors, including electroacoustic and psychoacoustic factors beyond the audiogram.
Methods: Thirty-five older adults with sensorineural loss completed audiometric evaluation and were tested, while aided, on QuickSIN under 12 conditions (high/low reverb × fixed/omni directionality × noise at 0°/90°/surround). Participants’ QuickSIN runs and hearing aid programs were replicated, recorded with KEMAR, and analyzed with HASPI_w2. [https://doi.org/10.1121/10.0017546] Participants completed five supra-threshold measurements with the iPAAS (iPad-based Psychoacoustic Assessment System).[https://doi.org/10.3390/audiolres11040061] Hearing aid benefits—spatial release from masking (SRM), directionality benefit (DB), and combined benefit (CB)—were derived independently from QuickSIN and HASPI scores. Group- and individual-level relationships between QuickSIN and HASPI benefit scores were analyzed. Linear models were fitted to quantify the variance explained by audiogram, HAPSI, and iPAAS.
Results: QuickSIN and HASPI scores correlated moderately overall (r = 0.61), but correlations between HA benefits derived from the two scores were low. However, both QuickSIN and HASPI captured the major acoustic manipulations at the group level: reverberation penalties, directionality gains, and 90° SRM. Stepwise linear modeling of HA benefits showed that audiogram explained up to 32% of DB/CB variance in low reverb, while iPAAS added up to 45% in high reverb; HASPI added < 2% at any step.
Conclusions: HASPI captured group-level differences, but had limited utility for individual-level prediction in this study. Supra-threshold abilities stood out in predicting HA benefits, which aligns with HASPI’s authors’ suggestion that “incorporating individual measurements of auditory function into the metrics could improve prediction accuracy”. [https://doi.org/10.1016/j.heares.2022.108608]
