Knowledge about how much speech-intelligibility improvement an individual hearing-aid user will achieve with specific settings is of great interest to fit a hearing aid optimally. While such knowledge exists for amplification, little is known so far about factors that determine speech-intelligibility benefit due to noise reduction (NR). This study aims to investigate to what extent real-ear measurements (REM), the audiogram, and spectro-temporal modulation detection predict NR-benefit.

Experienced hearing aid users were fitted with the same commercial high-end hearing aids using instant ear tips, verified with REM. NR-benefit was quantified as the improvement in speech reception threshold when changing from mild to strong NR in a spatial speech-in-noise setting. The Audible contrast threshold (ACT) test was used as a modulation-detection measure to quantify supra-threshold hearing deficits. Closedness of acoustic coupling was assessed using real-ear occluded insertion gain (REOIG) measurements.

The results show a high predictive value of REOIG for individual speech-intelligibility benefit due to NR, and the highest predictive accuracy for a linear combination of REOIG and ACT. The individual audiogram did not increase the predictive accuracy further and was a weaker predictor than the ACT-test.