Audiology, like other clinical sciences, strives to treat patients using consistent, evidence-based approaches. It is expected that caregiving professions ensure that all patients are managed using verified and generally accepted clinical procedures (Cox, 2006). As far as hearing aid fittings go, some aspects of common clinical practice have been extensively studied and well-described. Two clear examples of such are the use of multi-channel, non-linear amplification to map the full range of speech into the patient's remaining dynamic range, and the use of real-ear techniques to verify the acoustic effect of hearing aids. However, there are areas of clinical practice that have not been studied as extensively. Fitting hearing aids to patients who fall into one of the special populations, such as those with severe hearing loss, atypical audiograms, asymmetrical hearing loss and Ménière's disease, are just a few such areas that are not well defined.
Standard fitting algorithms - whether independent like DSL® m[i/o] (Scollie et al., 2005) and NAL-NL1 (Dillon, 1999), or proprietary to a manufacturer- typically were not designed to account for patients from the various special populations. They have been developed with classic hearing losses in mind: mild and moderate, flat and gently sloping symmetrical losses caused by aging and/or noise exposure, and accompanied by good speech recognition skills in quiet. For these classic losses, prescriptive approaches have served well over the years. The typical patient, after adjustment and fine tuning, usually ends up using gain and compression values reasonably close to the prescribed settings. Most clinicians will agree, however, that when fitting patients who do not fall within the classic hearing loss profile, the prescribed fitting parameters may not always work well.
This article proposes some alternative approaches to hearing aid fittings with special populations. It is important to note, as stated previously, that there is a general lack of large-scale clinical studies regarding hearing aid fittings with special populations. Further, the populations discussed in this article deserve significantly more coverage than is possible in one paper. With these caveats in mind, this article proposes some thoughts and considerations for alternative fitting approaches to be used with these often challenging clinical cases.
A General Approach to Fitting Amplification
All fitting algorithms are essentially driven by the patient's hearing thresholds. Although most fitting algorithms have a more sophisticated structure and philosophy, the major effect of prescribed fitting parameters is to correct for the patient's hearing loss. The poorer the hearing thresholds, the more gain is prescribed. For a patient with the classic flat to sloping, mild to moderate loss, this approach works well. It provides improved audibility in the mid to higher frequencies where there is important speech information, and although there is hearing loss, there is usually still a good amount of remaining hearing. However, the standard operational behavior of many professionals who fit hearing aids is focused on audiogram correction - that is, the more hearing loss there is, the more gain is required.
A reasonable analogy is to consider the way peg legs were fit to unlucky pirates of the past. The fitting approach was pretty simple: cut a sturdy piece of wood that was equal in length to how much leg was lost. To verify the fit, the pirate was stood up and checked to see if he tilted to the right or the left. If he tilted, a little fine tuning to the length of the peg leg was performed.
These days, modern prosthetic devices are fit in a more sophisticated manner (Wilson, 1998). Biomechanical engineers aim to integrate prostheses with the current anatomy and physiology of the patient, including the remaining bone, muscle and nervous tissue. It is not just about what the patient has lost, but far more important is what residual capabilities the patient brings to the task. The question is, "How can the prosthetic device best utilize the patient's current structures in order to provide the very best functional outcome?"
This philosophy also applies to the fitting of amplification. Remember, hearing is not altered by hearing aids;what is altered is only the sound that enters the auditory system. We simply cannot change how the patient's auditory system works. The task of the audiologist is to find the best combination of signal processing that makes optimal use of the remaining capabilities of the patient. Sounds are manipulated in such a way that the patient's auditory system can make best use of the modified signal. When dealing with patients who fall into the special population groups, it is essential to adopt this residual capabilities mindset. Standard fitting prescriptions will often fall short when viewed in relation to the remaining capabilities that the patient brings to the table.
When approaching any patient - especially those with some sort of atypical hearing loss - a wise strategy is to create a "theory of the hearing loss," similar to the "theory of the crime" so popular in forensic crime dramas. Develop a best educated guess of the status of the patient's auditory system based on the diagnostic data. Generate a reasonable estimate of what the patient can and cannot do with his or her hearing and then select a processing package that makes best use of the patient's capability. This provisional solution may or may not be similar to what an audiogram-based fitting rationale may call for. The task is not to match a target, but to make the most of the patient's auditory potential. Targets may or may not be a good guide, depending on the nature of the individual patient's auditory disorder.
Ski Slope Hearing Loss
This is the largest of the special populations. It includes patients with hearing within normal limits through at least 1000 Hz, with a rapid drop in sensitivity in the higher frequencies. The most common cause of this configuration is excessive noise exposure.
The best guidance when fitting these patients is nearly 30 years old. Skinner (1980) reported on the benefit of high frequency amplification for patients with ski slope losses at a time when few of these patients were being fit due to the limited bandwidth of hearing aids. However, she also pointed out that full high frequency audibility was not a reasonable goal due to comfort and sound quality problems. She stressed the importance of achieving a balance between what was provided in the mid and high frequencies.
The traditional audiogram-correction approach for this population calls for little or no gain except in the extreme high frequencies. However, these patients may not be able to make full use of high frequency audibility (Hogan & Turner, 1998;Ching, Dillon & Byrne, 1998). Subsequently, there has been a focus on determining the extent of true dead regions in the cochlea (Moore, 2003). Although there are ongoing discussions as to how accurately we can measure or predict the presence of dead regions (Hornsby & Dundas, 2009), the basic point remains: establishment of a pure-tone threshold does not guarantee that region of the inner ear has usable, high-quality hearing.
An alternative fitting approach for this group is to reduce gain in the extreme high frequencies, and instead, target a modest audibility enhancement in the mid-frequency transition region. Although these patients have the most loss in the regions above 2000 Hz, they have a partial hearing loss in the region between 1000 - 3000 Hz as well. This is also the area where the greatest amount of speech information falls (Kryter, 1962). Providing a modest gain enhancement in this region not only will boost speech audibility, it will also provide a better chance of avoiding the sound quality concerns often seen when an extreme amount of high frequency gain is used. A modest gain enhancement between 1000 - 3000 Hz will also allow for a more open fitting, reducing the chance of acoustic feedback and occlusion.
Patients with ski slope losses represent an opportunity to apply the approach of fitting to residual capabilities. The response of the device is focused on a region that works reasonably well for these patients, instead of a region that may be dead or unusable. Simply correcting for the amount of hearing loss does not take into account the quality of the remaining hearing in the extreme high frequencies nor the region where the greatest amount of speech information is present.
Patients with low frequency sensorineural hearing loss (SNHL) typically present with a moderate hearing loss below 2000 Hz, rising to normal or near normal in the higher frequencies. These losses are almost always congenital and are often genetically linked. Although Ménière's disease is a common cause of an acquired rising audiogram, it will be discussed under the topic of Medically Complex Hearing Loss.
A hearing loss correction approach for patients with low frequency hearing loss would call for significant amount of gain in the low frequencies and little or no gain in the higher frequencies. This approach enhances audibility in a region where there is minimal amount of speech information, and also greatly increases the risk of upper spread of masking. Further, these patients were historically fit with powerful low-pass hearing aids and closed ear molds in order to maintain the frequency response in the low frequencies. Before the widespread use a real-ear techniques for verification, patients fit in this manner would often end up with an insertion loss in the high frequencies.
Schum and Collins (1992) reported on a group of patients with low frequency SNHL who were fit with a variety of different amplification schemes. They found that the traditional approach of mirroring the audiogram did not work nearly as effectively as approaches that reduced gain in the lower frequencies and maintained at least a minimum amount of insertion gain in the higher frequencies. Again, when viewed from a residual capabilities standpoint, these patients are looking for amplification characteristics that allow them to hear the most important parts of the speech signal using regions of their hearing that seem to have the most residual capacity.
Using a residual capabilities approach, a minimum of 10 to 15 dB insertion gain at 2000 Hz and above would be recommended for these patients, even if hearing is normal. Insertion gain in the low and middle frequencies should be set at no more than 15 to 20 dB. This initial response should be evaluated subjectively by the patient, with primary focus on increasing or decreasing the low frequency gain based on the patient's judgment of loudness. Some patients may want more gain in the low frequencies, but usually they will prefer less. Once acceptable loudness of the low frequencies is achieved, attention can be paid to traditional fine tuning of the higher frequencies, with the goal being good speech clarity and acceptable sound quality.
Medically Complex Hearing Loss
The term "medically complex" as used here refers to patients with etiologies such as Ménière's disease, sudden SNHL, rapidly progressive losses, auditory nerve and other neurologically-related disorders, just to name a few. Medically complex etiologies cause disruptions and abnormalities within the auditory system beyond inner and outer hair cell loss. There may be changes in the mechanical or electro-chemical functionality in the inner ear, neural transmission abnormalities, excessive spread of masking, loss of complex across-frequency interactions, and other issues. In general, this category is a loose grouping of patients whose hearing losses represent a more recent, dramatic change in the auditory system, due to a significant insult causing widespread disruptions and abnormalities.
Some of the most common observations about patients with these etiologies are:
- Poor speech recognition, even in quiet at supra-threshold levels
- Speech recognition roll-over effects
- Fluctuations in either thresholds and/or speech recognition ability
- Restricted dynamic range
- Audible distortion
Unfortunately, some audiologic and otologic training materials have discouraged fitting amplification on medically complex patients with little clinical or scientific evidence presented to support those notions. Nonetheless, case examples exist which specifically demonstrate that these patients can be appropriately and satisfactorily fit. Feldman and Oviatt (1993) presented the case of an 84-year-old patient with a sloping mild-to-severe SNHL and good word recognition in the left ear, and a flat, severe SNHL with no word recognition in the right ear. The patient had a long history of hearing aid use on the left ear, but never used amplification on the right ear. The audiologists fit the right ear and gave the patient several weeks to adapt to amplification in the previously unaided ear. Upon follow-up, soundfield word recognition testing was performed in quiet and at +10 dB S/N. The patient scored 76% and 56% in quiet and noise, respectively, with the hearing aid only on the left ear. However, with the binaural fitting, word recognition scores improved to 92% and 88%. The additional hearing aid on the ear with no measurable speech understanding under headphones allowed a significant improvement in both quiet and noise. This case demonstrates the potential disservice to the patient which can occur based on arbitrary, exclusionary amplification rules. Current understanding of auditory processing in medically complex ears is so rudimentary that the only true test is often a trial hearing aid fitting. Hearing aid fittings are fully reversible with no lasting negative side effects. Therefore, it is difficult to justify not engaging a hearing aid trial in these medically complex cases.
Many patients who present with medically complex hearing losses have hearing thresholds in the severe range. Given this degree of hearing loss, the first inclination is to fit them with powerful linear hearing aids. The better solution for many patients is the use of advanced technology multi-channel, non-linear fittings. Medically complex patients with severe hearing loss often use significantly less gain than would be expected based on their hearing thresholds and often will only tolerate highly compressed sounds at low audibility levels.
Schum (1995) studied a group of 15 patients with a variety of medically complex disorders, including
Ménière's disease, progressive idiopathic losses and retro-cochlear disorders. The subjects were fit with nine different hearing aid processing strategies, including linear and non-linear approaches. Performance was evaluated using a panel of 18 different objective and subjective measures. The results showed that performance with multi-channel, non-linear strategies was significantly better than with linear approaches, with 11 of the 15 patients stating an overall preference for a multi-channel, non-linear option. A retrospective evaluation of the gain levels used by those 11 patients indicated that they typically used between 5 to 15 dB less gain for soft inputs in the mid to high frequencies than that prescribed by NAL-NL1.
The impact of medically complex hearing losses varies among individuals. Traditional fitting rules may not be anywhere near what that person's auditory system can optimally use. Thus, these patients require experimentation to determine how their remaining hearing can be most effectively utilized.
Severe Hearing Loss
Adult patients with long-standing severe SNHL typically have some history of using powerful linear amplification. Over the last decade, sufficient evidence has been collected (such as Flynn, Davis & Pogash, 2004;Souza & Bishop, 1999) that demonstrates that many patients within this category of hearing loss can benefit from multi-channel, non-linear amplification. The advantage of moving away from linear amplification for these patients is that non-linear strategies provide access to a greater amount of speech information. However, finding the right combination of signal processing options for these patients can be more challenging than with patients with lesser degrees of hearing loss.
Two factors account for the specific challenge when fitting these patients. First, they typically have an extensive history of amplification use and have become quite dependent on what their hearing aids can provide. In fact, clinicians over the years have commented on how patients with severe hearing loss can become attached to one make and model of hearing aid. Any attempt to update device technology may be met with resistance. Secondly, the variability from patient to patient increases in terms of how well the residual hearing can process complex sounds. Severe hearing loss can be the result of many different physiologic changes in the auditory system beyond simple inner and outer hair cell pathologies. There can be mechanical and metabolic disruptions, membrane disruptions, large sections of inner hair cell loss, neural cell death and disrupted across-frequency coordination. Simply measuring pure tone thresholds does not give a complete picture of the quality of the remaining hearing. As Boothroyd (1993) points out, an important factor when considering a patient with severe hearing loss is auditory resolution. Some ears simply cannot process complex inputs such as speech as effectively as others, again due to variations in hearing loss manifestation among individuals.
As indicated above, research over the last decade has indicated that there is a role for non-linear amplification for these patients. However, the approach to amplifying severe hearing losses needs to be adjusted to account for the potential for poor auditory resolution. Compression offers the opportunity to make a broad range of speech audible for patient with a reduced dynamic range. However, for those with severe hearing loss, the amount of compression necessary to make the full range of speech audible may degrade the signal so much that the ear cannot effectively extract information;this is not the case for all patients with severe hearing loss. As observed by Lamore, Verweij & Brocaar (1990), variability in speech understanding from patient to patient increases dramatically when moving from mild and moderate hearing loss into the severe range.
This increased patient to patient variability requires the audiologist to remain flexible when applying non-linear amplification. Some patients may be able to handle compression ratios on as high as 4:1 and still be able to extract information from the compressed signal. Others may need significantly less compression in order to perceptually decode the incoming signal. Invariably, fitting rationales that attempt to maximize audibility will call for very high compression ratios for those with severe hearing loss. If a patient trying such a strategy reports that speech sounds muddled or indistinct, that does not necessarily mean this patient cannot use non-linear amplification. The amount of compression used and, importantly, the speed of the compression may need to be adjusted. Slower acting compression approaches will maintain more of the linear aspect of the speech signal over shorter time windows. Therefore, a patient who seems to be reacting negatively to high amounts of fast acting compression may find that a slower acting approach works better.
There will be some patients who fall into the severe hearing loss range who simply cannot use high frequency regions effectively. For those patients, shifting the focus of the response of the devices to the regions below 1000 Hz may be necessary. Before giving up on amplification of a full range of speech inputs in the higher frequencies, it is important to experiment with differing amount of compression, and different speeds of compression.
The fitting process in general should be adjusted for patients with severe hearing loss. For patients with mild and moderate hearing loss, a traditional approach of fitting to prescribed settings and then fine-tuning may be appropriate. However, for those with severe hearing losses, the fitting process is likely to be more adaptive.
One of the first courses of action is to establish the upper limit of the patient's hearing. While many clinicians may not measure UCLs for the typical clinic patient, measuring UCLs ahead of time or assessing the highest acceptable setting of the MPO and loud gain controls in hearing aids is essential when working with patients with severe hearing loss. Since these patients will classically have a reduced dynamic range, having as much available headroom as possible is advantageous to the fitting. The ability to raise the maximum output of the device 10 dB may help to significantly reduce the amount of compression necessary.
Prescribed settings from the chosen fitting software are a reasonable place to start. However, given the concerns regarding using too much compression as discussed above, it is essential to evaluate the patient's ability to extract speech information from the compressed signal. In some cases, the patient will report straight away that speech sounds muddled or indistinct - a classic reaction from those who have a long history of linear hearing aid use. Other patients will have an opposite reaction, indicating a pleasant surprise at being able to hear more of the speech signal than they have for many years. If the patient reacts negatively to the prescribed settings, be careful not to react too quickly. Of course, you must ensure that the patient is willing to use the new devices. However, if the patient is moving from linear to non-linear amplification, there is likely a significant benefit available because of the new levels of audibility. It will be necessary to assertively manage the first days and weeks of hearing aid use to make sure that the patient doesn't reject the new devices simply because they sound different. If the patient truly cannot extract information from a compressed signal, then changes will need to be made. However, if it is just a matter of the sound being different, then strongly encourage the patient to give the new fitting some time. There is a very good chance that the benefit of improved audibility will become clear to the patient over the first days and weeks of use of the new devices.
From the residual capabilities viewpoint, the goal with fittings for patients with severe hearing loss is to find a way to package as much sound possible into the patient's remaining dynamic range and to use as much of the patient's remaining auditory capacity as possible. Older hearing aids, due to linear processing and limited bandwidth, may have not provided the patient with as much information as a patient could actually handle. Although adjustment to the new processing may take time, the goal is to get the most out of the remaining hearing that the patient brings to the task. Ultimately, by using a residual capabilities approach when fitting atypical hearing losses, patients will reap greater benefits and satisfaction from amplification.
The recent focus within our profession of applying structured and standardized approaches to hearing aid fittings is a positive and welcome step. Excellence in professional practice requires the audiologist to have a vast understanding of common and atypical audiologic presentations, as well as prescribed and alternative solutions. SNHL can take many different forms and simple rule-based fitting approaches will sometimes not be enough. Flexible and creative thinking on the part of the professional is essential in providing relevant and effective care for all patients.
Boothroyd, A. (1993). Profound deafness. In R. Tyler (Ed.), Cochlear Implants (pp. 1-34). San Diego: Singular Publishing.
Ching, T, Dillon, H. & Byrne, D. (1998). Speech recognition of hearing-impaired listeners: Predictions from audibility and the limited role of high-frequency amplification. Journal of the Acoustical Society of America, 103, 1128-1140.
Cox, R. (2006). Evidence-based practice in provision of amplification. Journal of the American Academy of Audiology, 16, 419-438.
Dllon, H. (1999). NAL-NL1: A new prescriptive fitting procedure for non-linear hearing aids. The Hearing Journal, 52(4), 10-16.
Feldman, A. & Oviatt, D. (1993). BiCros or binaural: A case study. Seminars in Hearing, 14(3), 229-233.
Flynn, M.C., Davis, P.B. and Pogash, R. (2004). Multiple-channel non-linear power hearing instruments for children with severe hearing impairment: Long term follow-up. International Journal of Audiology, 43, 479-485.
Hogan, C. & Turner, C. (1998). High frequency audibility: benefits for hearing impaired listeners. Journal of the Acoustical Society of America, 104, 432-441.
Hornsby B.W.Y., Dundas, J.A. (2009). Factors affecting outcomes on the TEN (SPL) test in adults with hearing loss. Journal of the American Academy of Audiology, 20, 251-263.
Kryter, K. (1962). Methods for the calculation and use of the articulation index. Journal of the Acoustical Society of America, 34, 1689-1697.
Lamore, P. J., Verweij, C., & Brocaar, M. P. (1990). Residual hearing capacity of severely hearing-impaired subjects. Acta Otolaryngologica Supplement, 469, 7-15.
Moore, B. (2003). An introduction to the psychology of hearing (5th ed.). San Diego: Academic Press.
Schum, D. (1995). Amplification options for severe distortional hearing loss. Final report to the U.S. Department of Education.
Schum, D. & Collins, M.J. (1992). Frequency response options for people with low-frequency sensorineural hearing loss. American Journal of Audiology, 1, 56-62.
Scollie, S., Seewald, R., Cornelisse, L., Moodie, S., Bagatto, M., Laurnagaray, D., Beaulac, S., & Pumford, J. (2005). The desired sensation level multistage input/output algorithm. Trends in Amplification, 9(4), 159-197.
Skinner, M. (1980). Speech intelligibility in noise-induced hearing loss: Effects of high-frequency compensation. Journal of the Acoustical Society of America, 67, 306-317.
Souza, P., & Bishop, R. (1999). Improving speech audibility with wide dynamic range compression in listeners with severe sensorineural loss. Ear and Hearing, 20(6), 461-470.
Wilson, A. (1998). A primer on limb prosthetics. Springfield, IL: Charles C. Thomas Publishing.