Dr. Ross on Hearing Loss
by Mark Ross, Ph.D.
For the past twenty-five years, many of the leading audiologists in the country have been meeting yearly at a conference in order to grapple with one or another audiological issue. This year, at the last such one that will be held (at least in its present form), the organizers chose as their theme, "Amplification: What do we really know?" According to the presenters, much of what audiologists do today is not well supported by evidence of user benefit. On what evidence, these audiologists asked, do average practitioners base the fitting procedures and the hearing aid recommendations that they make? Are their practices supported by the available scientific evidence? Will consumers be better served by a different hearing aid or hearing aid prescription? These are questions that physicians have been asking themselves in recent years, likely as a reaction to the increasing complexity of medical/surgical treatments. Now, perhaps as a sign of the growing maturation of the profession, the leaders in audiology are asking themselves the same questions.
According to Dr. Robyn Cox, a leading proponent of evidence-based practice, audiologists typically treat their clients in ways consistent with their personal opinions and backgrounds, but not necessarily on the basis of the best available evidence (which they may not be aware of). But, in a profession that prides itself on its scientific foundation, no procedure should be followed and no recommendation should be given without an awareness of the available relevant research literature. Optimally, every new hearing aid development, particularly those claiming to significantly improve a person's listening performance, ought to be accompanied by solid research evidence to support the marketing claims. But from I can see, this is more often than not the exception rather than the rule. Often, too, the evidence that is presented has been sponsored by manufacturers rather than by independent researchers. While this fact does not automatically negate the value of the research, it should at least raise a cautionary yellow flag when interpreting the results.
Actually, as Dr. Cox points out in an article in The Hearing Journal, there are hierarchies of evidence, with those on the bottom less valuable than those further up the scale. At the very bottom of the hierarchy lies "expert opinion," the level of evidence she identifies as the basis for most of the hearing aid recommendations now being made. Expert opinion can be quite useful -- indeed it is often the only kind of evidence that we have -- but it is not objective and it is not scientifically based. Consumers need more than some audiologist declaiming, "in my experience," as the primary rationale for a hearing aid recommendation. Because of their different backgrounds, expectations, and experiences, it is likely that audiologists' recommendations will differ among themselves. If expert opinions" differ for people with similar hearing losses, or indeed if the recommendations of various audiologists differ for the same person, then it does beg the question of who, if anyone, is right. Could a better, more scientifically-based, recommendation have been made? Perhaps a less expensive hearing aid would serve a person as well, or a different fitting prescription? These are the questions that an evidence-based audiology approach can help answer.
Falling at about the same level as expert opinion would be personal testimonials regarding a particular hearing aid or fitting, whether these are made in person or in the media. Testimonials are a very appealing marketing technique and a very common way that people are influenced to purchase specific hearing aids. When we hear of the marvelous results that one person has achieved with a certain new hearing aid, well of course we want the same for ourselves. It is a very human response. However, just because Mr. Jones is thrilled with a particular aid or fitting does not mean that Mrs. Smith will achieve the same results. We all come to the hearing aid experience with different histories, hearing losses, and expectations. The judgments of one person, while valid for that person, do not mean that someone else will perform similarly with the same type of hearing aid or fitting adjustments. While testimonials can be a factor in making hearing aid decisions, they should never serve as the sole or main criterion.
Often, too, the technical capability of a newly introduced hearing aid is offered as evidence for enhanced hearing performance. As with personal testimonials, this is a very effective and appealing marketing technique. The creative and perhaps unique manner in which some hearing aid processes incoming speech sounds is presented as prima facie evidence of superiority. Clever hearing aid engineers can do wonders with digital signal processing, and there is no question that modern hearing aids are technical marvels. Just about every issue of every trade journal seems to include a description of some notable advance. And, to be fair, I have no doubt that the technical performance glowingly described in the marketing blurbs will actually be found in a laboratory analysis. But the technical capabilities of a hearing aid, no matter how impressive they sound, no matter how logically they may appear to be related to improved hearing, are not proof of improved listening performance by a human being. This is a separate issue and has to be demonstrated in its own right.
What must be considered is that every hearing aid still delivers its sound into an impaired ear. At some point, the influence of the damaged auditory system is going to limit how well a person can hear with any sort of hearing instrument, no matter how sophisticated. This means that consumers may be spending a great deal of money on some "premium" hearing aid without obtaining significantly improved hearing performance over what they would get from less expensive hearing aids. Actually, I don't believe that any hearing instrument can fully compensate for normal biological hearing. (I'd love to be proven wrong, though.) The real-world challenge is to try to determine or estimate the maximum speech perception capability of an impaired ear and to try to get as close to that as possible with a hearing aid fitting, including, when necessary, advanced and costly hearing aid features.
From a research perspective, the most convincing type of study, the one that is at the top of the evidence hierarchy, is a design that employs randomized controlled trials. In this type of study, human variables are defined and controlled as much as possible. Test conditions are varied to preclude such possibilities as learning effects or fatigue from influencing the results. The type of subject is clearly described and results considered most valid when applied only to people with similar hearing losses; for example, studies of subjects with moderate high frequency hearing losses should not be interpreted as applicable to people with severe flat hearing losses. This implies that a study that lumps together all the clients seen at a particular center would be more difficult to interpret than one in which subject characteristics are more precisely defined (although still higher on the evidence hierarchy than expert opinion).
In the best type of study, expectations -- whether on the part of the subjects or unconscious biases on the part of the audiologists -- are controlled in the research design. An example is a double-blind design where the subject does not know which is the control hearing aid/condition and which is the experimental one. At the opposite extreme, we see studies in which a newly introduced hearing aid is identified as such to the subjects, who may be told (or led to believe) that they are wearing the "latest, most sophisticated instrument on the market today," whose performance is then compared to the subject's "plain vanilla" personal hearing aid. In this instance what would we predict the results to be? People expect and hope to hear better with the "latest" hearing aid. Not only don't they want to be disappointed, they don't want to disappoint the researcher who appears to have a stake in the results and who may implicitly or explicitly communicate the hoped-for results. Nobody is being mendacious or lying, just being human. Tester bias may be avoided in one of two ways: either by having the two aids/conditions physically indistinguishable from one another, where the testing audiologist does not know which is which; or by having two sets of audiologists involved, one to do the actual testing and the other to record the results.
As Dr. Cox points out, with the perspective of an evidence-based approach, one can ask pertinent (and long overdue) questions about many of the new features now included in hearing aids. For example, most modern hearing aids include more than one channel of amplification, ranging from two to ten or more. Does it make a difference how many channels a hearing aid includes? Is it a case of the more the merrier? Can people indeed hear better with ten channels as opposed to two, or three, or four? Do people with different hearing threshold configurations have different needs in terms of the optimal number of channels? Since a consumer might have to pay more for hearing aids with a greater number of channels, this not only becomes a question of performance but of cost.
Other features Dr. Cox identifies as requiring an evidence-based approach are: the effectiveness of different types of acoustic feedback suppression methods, the many variables involved with automatic gain control circuits, which fitting prescription to use, and the different types of directional microphones. And the list goes on. Of course the audiological research community has not been idle in investigating many of these areas. A great deal of evidence is already available and can be accessed through a perusal of the literature or an internet search.
For a good example of what evidence-based audiology is all about, consider the advent of cochlear implants when they were first introduced as single channel devices some twenty-five years ago. As a new and invasive procedure, they not only had to be proven to be safe but also to have clearly defined benefits. To do this required a rigorous and controlled multi-clinic study. Procedures were standardized between centers, and both objective and subjective responses were obtained from the subjects. This has been the pattern with the introduction of every new implant development, from the addition of channels, coding strategies to various kinds of electrodes. When multi-channel implants were developed, their effectiveness was compared to the previous generation of implants. Ditto with the coding strategies; as new ones were developed, their effectiveness was compared to the earlier ones, using exactly the same procedures and, often, the same test material and subjects.
Even now, after cochlear implants are well- established as a routine clinical procedure, implanted people are still systematically evaluated over time in a standardized manner. Follow-up evaluations are a routine component of the implant experience. And because of the standardized evaluation practice, the results obtained with any new implant feature can easily be compared to those obtained previously. From an evidence-based perspective, our experiences with cochlear implants should serve as an example for hearing aid assessments or, rather, non-assessments. In spite of the fact that hearing aids are more numerous than implants and have been with us much longer, their performance on individuals is still treated rather casually ("Is that better, Mrs. Jones?"). As one well-known comedian would put it, "They get no respect."
Suggesting that audiologists employ an evidence-based approach does not mean that they can ignore common sense and clinical insights. This is apparent if we consider the research results obtained with directional microphones. I can't think of a single controlled study with directional microphones that has failed to demonstrate improved speech perception scores relative to those obtained with omnidirectional microphones. However, when careful studies were done looking at the real-world use of hearing aids with directional microphones, where people had to rate both types and compare them, the data showing superiority of directional microphones were either weak or questionable. Thus, it would appear that the results obtained in both types of studies were mutually contradictory, even though both were based on excellent research.
But this is where common sense and clinical skills come in. Some directional microphones do not have a very high directionality index; while their benefits could be measured in a controlled situation, the acoustical and social situations of the real world are uncontrolled. In such situations, listeners could not be sure if they were hearing better (or worse) because of the effects of the directional microphones or because the noise level or the speaker had changed. The solution here is obvious: provide more effective (higher directionality index) directional microphones where the advantages would be both measurable and easily noticeable.
But even more important than the inevitable acoustical variations in the real-world was the realization that people simply did not know how to use directional microphone hearing aids in a way that could maximize their potential benefit. In this instance, the real-world research results, instead of casting a pall on the use of directional microphones, simply pointed out the need to help people learn how to maximize their effectiveness (for example, by ensuring that the major source of noise is to their rear). Unfortunately, helping people realize the full benefit from their hearing aids is often a forgotten component in the hearing aid dispensing process.
The most desirable type of research evidence is actually not a specific study at all, but a systematic review of a number of randomized controlled studies that have evaluated the same procedure. This review must be conducted by a professional who is a skilled researcher, aware of the many procedural problems that can affect the validity of the project, and whose interpretation of the results reflects the quality of the studies. While the findings of one excellent study can be quite convincing, three or four more on the same topic, all reporting similar findings, will be even more so.
There is no question that an evidence-based approach places a burden on practicing audiologists. Typically, they practice as they have been taught. Over time, as they acquire experience and build up their observations, they consider what has worked for them in the past and tend to continue with the same approach. As far as it goes, there is absolutely nothing wrong with this; it is the way that one becomes a skilled professional. But in a fast-changing field, the "truths" of one generation may not hold for the next. Things change, and a professional has the responsibility (and burden) of keeping up with evidence that challenges our previous modus operandi. Climbing out of a rut is not easy.
Hearing aid dispensers have to question themselves and their practices constantly. Are they basing their recommendation of a new hearing aid mainly on the technical descriptions supplied by the manufacturer and what the aid is supposed to do? Or is there high quality research evidence available that proves actual listening superiority? If not, or if they are unaware whether there is or not, then it is time for them to look further. They have to formulate a specific question relating to the hearing aid fitting, conduct a search of the available literature on the topic and, most importantly, be able to assess the quality of the evidence, which is something that they should have been trained to do. And then, after the recommendation has been made and followed, they have to be sure to evaluate the outcome. Is the hearing aid user indeed hearing better with the new hearing aid or processing strategy? Without an outcome analysis, the hearing aid dispenser may simply be repeating the same erroneous recommendation over and over, never really knowing if listening expectations have been fulfilled.
But it's not all science. An evidence-based approach still requires a clinician to relate to each individual client in a sensitive and understanding manner. People come into a dispenser's office with all kinds of issues and problems relating to their hearing loss. They may not want to be there at all, but have been pressured into the office by a spouse or adult child. They may have had a number of bad experiences before the current appointment. Research is not going to help much here; rather, what is necessary are the interpersonal skills of the practitioner and his/her ability to instill confidence in the client. And, of course, there is much that is not known, questions that the most diligent review of the literature cannot answer. So, yes, there are times when the judgment of the skilled practitioner is paramount - "expert opinion," if you will. But only after the evidence has been examined and appraised. In reality, the "science" and "art" of the hearing aid fitting process are inextricably intertwined.