gms | German Medical Science

Modern hearing aids (HA) offer many options to adjust gain settings and noise reduction algorithms to the individual users needs. While the prescription of gain to the HA user is well established using fitting rules like NAL-NL2 and DSL I/O v5, noise reduction (NR) settings appear to be a highly individual matter and standardized NR prescription approaches remain to be established. Audiologists and HA users thus have to apply a trial-and-error process to adjust the NR settings to a level that provides the right individual balance between speech intelligibility benefit and listening comfort. This study focuses on the speech intelligibility aspect of individual NR as well as brightness and soft gain (BSG) preference settings. Fifty-two bilaterally hearing-impaired experienced hearing aid users were fitted with Oticon More 1 receiver-in-the-ear (RITE) devices according to the NAL-NL2 fitting rule. Subjects were given two listening programs with identical gain settings but either different NR settings or different BSG settings. After an accommodation period of two to four weeks, subjects participated in six field periods of approximately four weeks each and were asked to compare and report on the two listening programs in different listening situations using either paper or online questionnaires. During the first four field periods, NR settings were systematically varied to find the NR setting that was individually preferred by the HA user while the BSG setting remained on the preset level (in line with the NAL-NL2 target curve). In field periods five and six, NR settings in both listening programs remained at the subjects final setting after field period four, while BSG settings were adjusted. After completing all six field periods, subjects ended up with their final NR and BSG setting. Speech reception thresholds (SRTs) were measured in the final NR and BSG setting, in the final BSG setting with strong NR, and in the preset BSG setting with strong NR. SRTs were assessed using the German hearing in noise test (HINT) in a spatial scenario with frontal speech and symmetrically placed distracting talkers. Results showed a strong impact of the NR settings (p < 0.001), dividing subjects into two groups that either gained (n = 10) or lost (n = 42) speech intelligibility in the final NR setting in contrast to the strongest NR setting. No significant difference was found between the final and prescribed BSG settings (p = 0.700), but again subjects fell into two groups that either benefited (n = 24) or obtained a disadvantage (n = 28) from the preferred BSG setting. Factors that may drive the individual speech intelligibility sacrifice in favor of the final setting will be discussed. This study was funded by the William Demant Foundation.