Relation Between Self-Reported Outcomes and Real-Ear Insertion Gain as a Function of Generic Fitting Prescriptions

Background
It is essential to understand the relation between the individual real-ear insertion gain (REIG) and self-reported outcome measures. A large cohort study can provide insight into the efficacy of current practices in hearing aid (HA) rehabilitation according to the accuracy of the fitted gain.
Methods and Materials
A total of 1231 participants with a valid REIG for both ears at 65 dB input level having answered all the questions in the Danish version of the IOI-HA and the SSQ administered in the BEAR project were included in the study. The participants included both first-time (n = 895) and experienced HA users. The differences in gain from individual REIG to that prescribed by four generic gain prescriptions, namely, NAL-NL2, NAL-RP, one-third gain (OTG) rule, and the half-gain (HG) rule was calculated. Clustering of these individual gain differences to the four generic prescriptions was performed using the k-means method. The resulting cluster was then used as a single categorical unit representing the gain prescribed to the individual HA user. Multiple linear regression analysis was performed for the SSQ data, and an ordinal regression analysis was used for the IOI-HA data.
Results
The k-means clustering resulted in two significant clusters characterized by gain differences at higher frequencies. The analysis showed that patients closely fitted to NAL-NL2, NAL-RP, and OTG reported higher IOI-HA Factor 1 scores (introspection about the HA) than the cluster of patients fitted away from these gain prescriptions. The multiple linear regression of SSQ showed a contrasting result for gain differences to OTG, where a better SSQ speech domain was reported by patients that were fitted away from OTG prescription. A significantly higher SSQ speech domain score was reported for patients with a higher gain difference to HG, analogous to OTG. Further analysis showed that this effect originated from the gain difference at 8000 Hz between the clusters. If excluded, the cluster was no longer a significant predictor of SSQ speech domain score for the gain difference to OTG and HG prescriptions.
Conclusion
The findings from the study suggest a correlation between the actual fitted gain and self-reported outcome.