Fine-Tuning in Modern Hearing Aids

The precalculated gain settings of a hearing aid are essential for the fitting process. When only a little fine-tuning is required, the hearing care professional can concentrate on more advanced aspects of fitting and counseling. In this retrospective field study, 1028 real-life fittings of a modern hearing aid from 69 dispensing professionals were analyzed to determine the amount of fine-tuning which is actually done in practice. In general, the results show only a little fine-tuning was needed, and that fine-tuning effort is affected by the amount of hearing loss.

As with most products, first impressions are important, and spontaneous acceptance is a key factor for successful hearing aid fittings. The accuracy of the precalculated hearing aid settings plays a fundamental role in achieving spontaneous acceptance.

When only minimal adjustments are required to satisfy a client’s first impression of sound quality, loudness and speech understanding, then a positive and solid basis is established for the remainder of the fitting process. This allows the hearing care professional to focus on more advanced aspects of the fitting and counseling process to improve the long-term client satisfaction. These more sophisticated aspects may include adaptive directionality, multi-program access, or automatic program selection, to name only a few.

In contrast, if the precalculated gain settings differ considerably from the desired settings, initial disappointment with the new hearing instruments can overshadow all other subsequent aspects of the fitting.

The precalculated settings might be derived from prescriptive gain formulae such as DSL or NAL, or from manufacturer’s proprietary rules. But how appropriate are the precalculated gain settings of modern hearing instruments? Do they precisely meet the individual amplification needs of the clients? In other words, how much fine-tuning of the precalculated gain settings is actually done in practice?

Examination of 1000+ Fittings
To assess the quality and precision of the precalculated gain settings in a modern hearing aid (Phonak Perseo), a large number of real-life fittings were scrutinized. In total, 1028 fittings (1663 ears) from 69 hearing care professionals were retrospectively analyzed. The hearing losses ranged from mild to profound, with a PTA of 51 dB. The mean age of the clients was 69 years (SD ±15 years). In 96% of all fittings, the proprietary “Phonak Digital” fitting rule was used (Phonak digital ski slope: 3%; DSL[i/o] or NAL-NL1: 1%). All available Perseo models were used.

Figure 1. Average fine tuning (ie, gain deviations to precalculated settings).

The average gain adjustment was found to be -0.6 dB. The extent to which gain adjustments were required is illustrated in Figure 1. Shown is the mean gain deviation across all 20 frequency bands between the precalculated gain and the gain after fine tuning. The smaller the deviation, the better the initial fit. The error bars show ±1 SD across subjects. The average gain deviations are quite small across frequencies (-0.6 dB). There is only a slight trend towards decreasing the initial gain settings in the high frequencies. For most frequencies, the deviations fluctuate around zero. This indicates that the initial fit is indeed appropriate and does not systematically provide too much or too little gain.

Figure 2. Individual adjustments of precalculated gain settings (averaged across frequency) related to the respective hearing loss. Each data point represents one client.

Fine-Tuning and the Degree of Hearing Loss
The relationship between individual hearing loss and related fine tuning is illustrated in Figure 2, which shows a scatter plot of average hearing loss versus average gain adjustments.

For mild hearing losses, there is a trend to slightly reduce gain (negative values) on the initial precalculated gain settings. It should be noted that the proportion of first-time users in this segment is likely to be higher compared to the more severe losses. These clients are not used to amplification and often ask for gain reductions, especially in the high frequencies.

Figure 3. Standard deviation of gain adjustments for different average hearing losses. The higher the standard deviation, the more distinct fine-tuning actions were taken in terms of decreasing or increasing the gain.

Figure 3 shows the variability (ie, the standard deviation of gain adjustments) across subjects for different hearing losses. The higher the standard deviation, the more distinct the fine-tuning actions are in terms of increasing or decreasing the gain.

The results show that the greater the hearing loss, the higher the observed variability in gain adjustments (ie, more fine-tuning had to be done). The individual’s “window of tolerance” for amplification narrows with increasing hearing loss, thus requiring more gain adjustments. This corresponds with clinical experience: the more pronounced the client’s hearing loss, the more fine-tuning is generally necessary.

Overall, with an average standard deviation of less than 5 dB, the variability across subjects is rather small. In 32% of all fittings, no gain adjustments had to be done at all.

In summary, the precalculated gain settings did indeed correspond well with the settings which were actually chosen in practice. Thus, it can be concluded that the average initial acceptance of the precalculated gain settings was quite high. This allows the hearing care professional to focus on more advanced aspects of fitting and counseling, which in turn will increase long-term satisfaction and the overall fitting success. The data reported in this field study is confined to one hearing instrument brand and model, as well as (primarily) a proprietary fitting strategy; obviously, it would be of interest to analyze the wide range of hearing aids of different technologies, manufacturers, product lines, and fitting strategies to assess required adjustments on a more industry-wide basis. w

Acknowledgements
The author thanks Olivier Domken, Katja Eckhardt, Iris Arweiler, and Danny Bürkli for their support in collecting and analyzing the data.