Clinical probe-microphone measurements and cellular telephones were both introduced in 1979. Care to guess which technology has been more widely accepted by professionals dispensing hearing aids?
Yes, it was 26 years ago that Earl Harford, PhD, described a “clinically friendly” method of measuring the output of hearing aids in the real ear. The version of the technique he described had been used in the laboratory as early as the 1940s. I say “clinically friendly,” but I doubt that patients rated Earl's method as friendly. Unlike today's probe-tube microphone systems, this early method of real-ear measurement involved putting the entire microphone (about 4 mm by 5 mm by 2 mm) down the patient's ear canal.
In the early 1980s, the first computerized probe-microphone system, the Rastronics CCI-10 (developed in Denmark by Steen Rasmussen), entered the U.S. market. This system had a silicone tube (which was part of the calibration process) attached to the microphone, thus eliminating the need to place the microphone in the ear canal.
By early 1985, three or four manufacturers had introduced computerized probe-microphone equipment, and, it seemed, hearing aid fitting and verification would be changed forever. A lot has happened since 1979, and today most dispensers consider their cell phones essential. However, when it comes to being fitted with hearing aids, most patients still don't receive probe-mic verification of gain and output. Why is this?
USE OF PROBE-MIC MEASURES
The Hearing Journal has published several Cover Stories on probe-microphone measures going back to the 1980s. We recognize that these measures have many uses. For example, probe-mic measures, such as the real-ear coupler difference, are essential for fitting infants and children. This report, though, will focus primarily on the use of probe-mic testing for verification of hearing aid performance with adults.
Since the early 1970s, it's been common to conduct surveys on the procedures practitioners use to fit hearing aids. If you're old enough, you might remember when a survey told us, “Eighty-eight percent of audiologists use speech ‘discrimination’ to pick the best hearing aid.” HJ conducts surveys of this type, as do other journals and individual researchers. And, when it comes to probe-mic measures, their findings agree.
Surveys generally show that about a third of audiologists/hearing instrument specialists conduct probe-mic measures most of the time, another third use them some of the time, and the remaining third seldom or never use probe-mic assessment.
There are probably several reasons why probe-mic measures have never become routine practice. Let's examine some of the most commonly mentioned reasons.
Uncertain correlation with hearing aid satisfaction
The simple act of conducting probe-mic measures, or even the careful matching of hearing aid output to prescribed targets does not always have a profound effect on overall patient satisfaction. As Larry Humes, PhD, and others have shown in their research, satisfaction with hearing aids involves many complex factors. If we suddenly discovered a 1:1 relationship between the output of hearing aids in the ear canal and patient “happiness,” probe-mic equipment would be flying out of storerooms and into clinics. Unfortunately, creating happiness is not that simple. Getting the ear canal output right is only one component of the overall fitting, perhaps too subtle a component for many dispensers to appreciate.
Complexity of modern hearing aids
When WDRC was introduced, people said, “You can't do probe-mic with WDRC hearing aids.” Then along came multichannel programmable and people said, “You can't do probe-mic with programmable hearing aids.” Most recently, the naysayers have told us: “You can't do probe-mic with digital hearing aids.”
As most of you know, all this is simply untrue. However, if a dispenser isn't too keen on conducting real-ear verification in the first place, it doesn't take much of a push to convince him or her to leave that dust cover firmly attached.
Regarding today's hearing aids, I would argue just the opposite of this “Team Can't Do” credo. That is, the more complex the processing of the hearing aid and the more automatic, adjustable, and adaptive features it has, the more important it is to determine how this hearing aid is functioning in the real ear for different inputs.
Bill Cole, president of Audioscan, a leading manufacturer of probe-mic equipment for 20 years, said: “Dispensers need to know that it is not only possible to use modern probe-mic systems to fit digital hearing instruments, it is almost mandatory because of the complexity of these instruments and the software that programs them. They need to realize they cannot rely on manufacturers' ‘first fit’ algorithms to deliver the potential of the instruments they fit, nor can they rely on manufacturer's software simulations to show them what is going on in their client's ear.”
Today, nearly all hearing aid fitting software provides graphics showing simulated gain and output for different inputs and, in some cases, a comparison with prescriptive fitting targets. These simulated graphics often are available for 2-cc coupler, real-ear insertion gain, and/or the real-ear aided response. They are useful for making hearing aid adjustments. For example, if you choose to boost gain in the highs by 5 dB, it's easy to verify this on the fitting screen.
Unfortunately, as alluded to by Bill Cole, some dispensers see these graphics as a substitute for probe-mic verification. They assume that what's on the screen is similar to what's being presented in the ear canal. This is a risky assumption, even for hearing aids with basic signal processing, and it becomes riskier still when digital noise reduction and other advanced signal processing are involved. But, it probably has contributed to the apparently common belief that gain and output need not be verified in the real ear.
Poor training programs
It is sometimes suggested that professional training programs fail to emphasize the need for real-ear verification, and therefore a generation of students has graduated who believe it is unimportant.
But, if this was true in the past, it certainly doesn't seem to be today. At a recent “teaching hearing aids” conference at the University of Pittsburgh that drew representatives from more than 30 AuD programs, the clear consensus was that probe-mic measures were an essential part of the hearing aid curriculum. Of course, what is taught in the classroom is not always practiced when students walk down the hall to the clinic or leave the university.
As you may recall, one reason given for expanding audiology training to 4 years was the need to provide more training in hearing aids. And the plan has worked. Programs that used to have one hearing aid class now have two or three.
We thought it might be interesting to look at the most recent Hearing Journal /Audiology Online survey results, conducted earlier this year, to see if this additional training has influenced how students verify hearing aid performance. We looked at four groups of audiologists who dispense hearing aids: young AuDs (graduates of residential programs with less than 5 years' experience); young master's degree holders (with less than 5 years' experience); older AuDs (distance-learning graduates with over 15 years' experience); and older master's audiologists (over 15 years' experience).
What did we find? When we calculated “routine use of probe-mic measures,” three of the groups were essentially the same (38%-42% routinely use): young AuDs, young master's, and older AuDs. But, among the older master's audiologists, only 21% reported routinely using probe-mic verification. This last group makes up the largest portion of audiologists fitting hearing aids (and of survey respondents), which is why when all four groups are added together we end up with a one-third use rate.
That there is no difference between new AuDs and new master's audiologists probably doesn't surprise you. The difference between the two older groups, however, is interesting. You could interpret this to mean that the AuD group's distance-learning hearing aid class prompted them to “see the light” (or “tube”). Or, it could simply be that these were the people who already were doing probe-mic testing before they enrolled.
Why is it that so many people who have been taught the importance of hearing aid verification in school decide that this testing is unnecessary when they start fitting hearing aids in the real world? If anyone knows the answer to that question, it's probably Michael Valente, PhD, of Washington University in St. Louis, who has been teaching graduate students about probe-mic measures since they were introduced. To Mike, fitting hearing aids without probe-mic verification is worse than wearing white shoes after Labor Day! When I asked him why students change after they leave the university, here's what he said:
“There are probably several reasons, but first, according to the clinic coordinator at one graduate program, in 16 of 20 external clinical sites where she sends her students for clinical experience, probe-mic equipment is never used or used only in ‘special cases.’ Therefore, some students wonder if probe-mic verification really is necessary if seasoned audiologists with successful practices do not use it.
“Second, I believe a majority of students graduate understanding the need for probe-mic measures to implement a ‘best practice,’ but the facility in which they are employed either does not have the equipment and/or has staff who do not promote its routine use. It is very difficult, if not impossible, for new graduates to arrive at their first job and change the method of practice.”
Cost of the equipment
Probe-microphone equipment would be considered a major purchase by most dispensers or clinics, and it's tempting to think its cost is a major factor holding back the use of probe-microphone verification. That notion, however, loses steam when you consider the results of our survey of 6 years ago, when we limited our tally of probe-mic use to practitioners who owned the equipment or had access to it. In this group, routine use was still only 42% and, most significantly, 45% of dispensers who owned probe-microphone equipment said they seldom or never conducted real-ear procedures.
Also arguing against cost as a factor is that, from a “return-on-investment” standpoint for dispensers selling hearing aids, probe-microphone equipment would seem easier to justify than equipment used for diagnostic testing.
Finally, for the budget-minded, there are many lower-cost models and good used systems. To put things in 2005 terms that we can all understand, on the 4-year plan you can acquire one of these systems by making payments approximately equal to the cost of ordering a Starbuck's double-tall latte every day, something that appears to be affordable even for graduate students.
A final factor that may lead to the non-use of probe-mic measures might be the overall dedication of individual professionals. This is a sensitive topic, so I thought I'd let George Frye handle it. As president of Frye Electronics, Inc., George has been building hearing aid assessment equipment for 32 years. Here's his take on the dedication issue:
“The population of people who fit hearing instruments includes those very dedicated to the task, those who are not too aware, and everyone in between. Those who are most dedicated are going to use every tool at their disposal to make sure the hearing instrument is properly fitted and working well. Those on the other end of the dedication scale probably use an audiometer just because they are required to.”
SUPPORT FOR PROBE-MIC MEASURES
So far, we've talked mostly about why probe-mic testing is not routinely conducted, and we've only implied that using it to verify hearing aid performance is a good idea. So, now let's briefly review the intended role of these measures in the overall fitting process. Many different measures can be made when a probe microphone is placed near the eardrum. These measures, some of them from an ANSI Standard (S3.46), were discussed in detail in our last Cover Story (March 1998) on this topic, so we won't wade through them all again.
While probe-microphone measures are useful for measuring the real-ear effectiveness of specific hearing aid features (e.g., directional technology, digital noise reduction, telecoil function), assessing the occlusion effect, and establishing correction factors (e.g., the RECD), their primary purpose is to assess gain and output for the settings that the patient will actually use for listening to speech. That is, the measure answers the fitting question, “Are the hearing aids delivering the appropriate output for soft, average, and loud speech signals, and, are high inputs delivered at a comfortable and safe level?”
Probe-microphone measures are not a way to fit hearing aids. Rather, they verify whether or not the hearing aids are performing in the manner that the person who fitted them believes they should be performing. The dispenser's goal may be as simple as “to make sure that soft speech is audible and that loud speech is loud but not too loud.” Or, the verification process could involve using specific fitting targets for different inputs from validated prescriptive methods, such as those provided by the NAL-NL1 and DSL4.1.
Audiologists working in the Veterans Administration (VA) have a huge impact on hearing aid fitting practices, both because they fit nearly 15% of all hearing aids dispensed in the U.S., and also because, at some point in their training, most AuD students work at a VA clinic. I asked Lucille Beck, PhD, the VA's national director of audiology and speech pathology services, for her thoughts on the use of probe-mic verification. Dr. Beck, who has worked with several groups in establishing hearing aid fitting guidelines, replied:
“There are many components of a responsible hearing aid fitting [and] determining and verifying the appropriate frequency gain and signal processing characteristics in the ear for every patient are essential. This testing is part of audiology's professional responsibility. As the VA works toward adopting standard outcome measures, probe-microphone measures will be incorporated into the mix.”
Indeed, probe-mic measures have been part of every published hearing aid fitting protocol since they were introduced. The most commonly used hearing aid fitting protocol was published in the American Journal of Audiology in 1998, written by five noted audiologists: Drs. Ruth Bentler, Richard Seewald, Tim Trine, Mike Valente, and Dennis Van Vliet.
For some, adherence to the protocol simply is “business as usual.” Therese Walden, AuD, head of the audiology clinic at Walter Reed Army Hospital where probe-microphone measures have been used for 20 years, states: “For us, probe-mic testing is routine. With the ability we have today to evaluate hearing aids with different input signals, using probe-mic measures we can set prescriptions that address the patient's needs. In addition, we use real-ear output to determine directional-microphone function and determine the effectiveness of digital noise reduction on average level input signals.”
There certainly are compelling reasons to include probe-mic measures in your fitting protocol. We know from several reports that manufacturers' default fittings are often quite different from the gain and output of validated methods and that frequently the simulated real-ear gain shown in the fitting software differs significantly from what is present in the real ear. Given these two facts, if a dispenser is concerned about aided audibility, maximizing intelligibility, and listening comfort for different input levels, there is no alternative to probe-microphone measures.
There are a couple of trends emerging in probe-mic measurement, and new equipment and test signals are allowing us to do old things more efficiently, accurately, and reliably.
First, in general, more people seem to be using the REAR rather than the REIG for hearing aid verification. Observing the hearing loss and dynamic range referenced to ear canal SPL is not a new concept. Norm Erber and others were talking about this in the 1970s, and this approach always has been used by the DSL, introduced over 20 years ago. Historically, however, dispensers have preferred the REIG. This is understandable, as in the early days of probe-mic testing, nearly all the prescriptive methods were verified based on gain (e.g., Berger, NAL, Libby, POGO). Plus, in concept, the REIG (or REIR, as it was called back then) was a measure very similar to functional gain, which was the training ground for early probe-mic adopters.
For children, I believe that most would agree that the REAR is the preferred verification method. For adults, you could argue that the REIG remains a reasonable method to verify hearing aid performance. Regardless, more people seem to be using the REAR with adults, which to some extent is related to the second trend.
Using speech/speech-like signals
Probably the most significant change in probe-mic testing in the past few years has been the increased use of input signals that are speech or speech-like. Using real speech or speech-like signals to evaluate real-ear hearing aid performance has been discussed (and done) for decades. Today, the prevalence of hearing aids with multiple channels of compression and digital noise reduction has made it more of a necessity than a luxury.
A few years back on Page Ten of the Journal, I wrote that it was okay to conduct your probe-mic testing using old-style signals with modern digital hearing aids if you simply turned off the noise reduction when you conducted the measure. I was wrong. I had been led to believe that when speech was the primary signal, the output of the hearing aid would be the same whether noise reduction was on or off, and, therefore, testing with it off would tell you the output for speech with it on, which is what you need to know to program the hearing aid correctly.
Well, as Ruth Bentler and others have pointed out, based on their laboratory testing of several hearing aids from different manufacturers, it does work that way for some products. However, for others, there is a significant effect on the speech signal when noise reduction is turned on. Assuming your patients have noise reduction turned on for their primary listening program, this means we need to do real-ear testing with noise reduction on, and use a test signal like speech, or a signal that the hearing aid thinks is speech.
Fortunately, manufacturers of probe-mic equipment were well aware of this situation, and today we have many appropriate signals to choose from. In some cases, this is a speech-shaped signal with modulations; in others, it is a recorded real-speech signal. And some equipment provides the option of using live speech. When real speech is used, either recorded or live, the hearing aid output is commonly displayed in the REAR mode, and often the procedure is referred to as speech mapping.
Does the use of real-ear “speech mapping” improve the hearing aid fitting? To some extent, this depends on what procedure it's replacing. If you previously were relying on the simulated curves from the fitting screen, then you have some surprises in store. On the other hand, if you typically fitted an “average adult” by carefully matching your NAL-NL1 REIG targets for soft, average, and loud input signals, my guess is your speech mapping will look equally good (assuming you don't create different targets).
If your REIG showed little or no gain above 2000 Hz, then the same problem would be revealed with speech mapping. The obvious benefit of real speech is that you can see the effects of compression and other signal processing that otherwise may not be apparent.
Several anecdotal reports and at least one research study suggest that hearing aid patients prefer being “speech mapped” than having traditional probe-mic measures done to the extent that “mapping” creates greater satisfaction with hearing aids and reduces return office visits during the first few weeks of hearing aid use. I'm only guessing, but it could be that seeing the amplified speech spectrum placed within their dynamic range assures patients that all is well and makes them more willing to let acclimatization take its course. It also could be that dispensers become better counselors when they conduct speech mapping, as the screen graphics provide several talking points.
Real speech and validated methods?
As mentioned above, when it comes to probe-mic testing, the buzz seems to concern the use of speech mapping. In some cases, however, it seems that the excitement over real-ear real-speech testing has led dispensers away from validated prescriptive fitting methods. But, why not use the latest method of probe-mic verification without abandoning established fitting targets?
Before making any recommendations on this, I checked in with experts on our two most established and validated methods, the DSL4.1 and the NAL-NL1. Sharing their thoughts on the topic are Susan Scollie, PhD, assistant professor at the University of Western Ontario, and Harvey Dillon, PhD, director of research at the National Acoustic Laboratories (NAL) in Australia.
I first asked if the current trend of conducting real-speech verification was compatible with their respective fitting algorithms.
Yes, for the DSL it fits very well. Our speech targets are meant for comparison with long-term levels of speech, in 1/3-octave bands. So, if the speech-based verification system provides this type of analysis, it is actually a very nice match between the philosophy of DSL's targets (to provide good audibility for speech) and what is actually measured (speech).
The NAL-NL1 prescription is based on maximizing speech intelligibility while controlling loudness. The derivation of the procedure explicitly uses the long-term average speech spectrum based on a multi-language international study. Consequently, it is entirely reasonable to use actual speech as a test signal to see how good a job a hearing aid is doing at giving the prescribed amplification. A major proviso is that the comparison between actual performance and prescription has to be done in a valid way. But, fortunately, there are several possibilities.
I also asked our prescription fitting experts about the use of recorded versus “live” speech input signals.
My preference for use with the DSL is recorded speech. Live speech is a handy extra, but not for routine use for reasons of replicability, level control, and clinician fatigue.
Recorded speech is preferred because the equipment manufacturer can ensure that the long-term spectrum of the test signal is a reasonable approximation of the international average spectrum used to derive the NAL prescription.
So there you have it: all good news concerning the use of validated methods and real-speech real-ear testing.
WILL USE PATTERNS CHANGE?
So far we've determined that probe-mic measures are an essential component of the hearing aid fitting process, that protocols and opinion leaders strongly support their use, that we now have appropriate signals and analysis to evaluate the signal processing of modern instruments in the real ear effectively, and that these measures fit nicely with our validated prescriptive methods. What a perfect world! Does this mean dispensers will soon start conducting probe-mic measures more frequently?
Some see the recent interest in using speech signals as sparking increased use. Bill Cole does, as he states: “We have seen a large increase in interest since the introduction of our speech-based system. This interest is coming from private practices as well as agencies and institutions. We sense that many dispensers have been frustrated with their inability to objectively verify their digital instrument fittings and, now that it is feasible, they are moving to take back control of the fitting process.”
Will new guidelines make a difference? Mike Valente chaired the committee that recently wrote the new evidenced-based hearing aid fitting guidelines, soon to be published by the American Academy of Audiology. Here's what he said:
“I do not for one moment believe that the adoption of this guideline will have any more impact on how hearing aids are dispensed than the one I chaired for ASHA back in the mid-1990s. Nothing really changed after that was published, and I suspect little will change after the current guideline is published.”
Regarding increased compliance with protocols, Valente added: “I do not believe that probe-mic and other methods of verification will be part of the ‘routine practice’ of dispensing hearing aids until there is some form of legislation or mandates from third-party payers that reimbursement for hearing aids will not occur unless there is documentation that some method of verification of the fitting has been performed.”
On a similar note, George Frye stated: “If the fitting process were to be unbundled, with separate charges made for tests as well as for the basic hearing instrument, then the addition of a real-ear check would have a cash incentive. We are starting to see more third-party pay systems in hearing instrument dispensing. I predict that most of those who will be reimbursing dispensers for delivery of hearing instruments will be demanding written documentation of the fittings.”
It would be nice to believe that verification need not be driven by reimbursement and that dispensers would conduct probe-mic testing simply because they knew it made a difference in the fitting. Yet, given that neither of these factors seems to be moving the majority, perhaps the “Big Brother” approach will increase probe-mic testing in some arenas.
It's tempting to think of government when Big Brother comes up, so I asked Lu Beck if the VA monitors compliance with recommended protocols. She replied: “The VA has adopted the practice guidelines supporting probe-microphone verification for use as our protocol and standard of care, although every clinic is responsible for its implementation and reports to its own leadership. We do not have a national mechanism for checking compliance and no doubt there is variability from clinic to clinic.”
The Big Brothers in the lives of dispensers in private practice may be state laws, possibly third-party payers, or perhaps a professional organization's code of ethics. But none of these relate very specifically to how hearing aid fittings are verified. However, what is becoming more common is for offices to become part of a larger dispensing network, which may add a degree of “Big Brotherness.”
Two of the largest dispensing networks in the U.S. are the Amplifon Group (Sonus, Miracle-Ear, National Hearing Centers) and HearUSA. I asked Brian Taylor, senior manager of professional development for Amplifon, if that company had formal policies regarding probe-mic testing.
He replied: “We invest a significant amount of money each year in the purchase of new probe-mic equipment for our offices and we do extensive training. This year alone we are sponsoring a dozen workshops on real-ear verification. We do all we can to ensure that our providers use it, and that they understand that when used appropriately, this practice will contribute ultimately to higher patient satisfaction and benefit.”
Dennis Van Vliet, AuD, vice-president of professional services at HearUSA, presented a similar, but stronger viewpoint: “Using probe-mic verification and making hearing aid adjustments based on these findings to ensure that the correct output is present for different inputs is part of our standard protocol. We do mandate it, unless there is an insurmountable problem, in which case we'll accept soundfield thresholds as a very poor substitute. All our audiologists undergo chart peer review by the quality assurance supervisors and are held accountable for meeting our standard of care. If they do not follow protocols, they are enlisted in a more stringent peer review process, and are subject to disciplinary action if they do not comply.”
Finally, I asked Therese Walden of Walter Reed what would happen if she learned that one of her staff was not using probe-microphone verification. She replied: “We require real-ear testing for all hearing aid fittings. If I ever had a staff member who didn't follow this protocol, I'd get my Uncle Guido after them.”
Maybe the “Big Uncle” approach is the solution. Are there enough Uncle Guidos to go around?
SCHWEITZER PROVES NIELS BOHR WAS RIGHT
Back in 1986, in HJ's first Cover Story on probe-mic, one of the audiologists interviewed was Chris Schweitzer, PhD, who provided us with this bold prediction: “I see a real readiness for real-ear systems among dispensers. I think that 80% of them will rely on real-ear fitting measurements in 5 years.” Chris operates a private dispensing practice in Colorado, and I caught up with him recently.
Mueller: So, Chris, what exactly were you thinking back in 1986 when you said that 80% of dispensers would be using real-ear measures in 5 years?
Schweitzer: I said what? I'd like to deny it, but you're the kind of guy who saves his 1986 Hearing Journals, so let me fall back on a Niels Bohr quote: “Prediction is very difficult, especially if it's about the future.”
Mueller: It sounds as if you were pretty excited about probe-microphone measures back in 1986.
Schweitzer: I was, and still am. Back then, a lot of really smart folks were exploring and exploiting the new method of “in situ” hearing aid performance. Somehow, I got in the middle of it and ended up chairing a newly formed ANSI working group charged with trying to standardize the protocols and educating clinicians. Everyone was pretty excited about the prospect of reducing consumer complaints by individually characterizing the hearing aid properties as delivered at the eardrum.
Mueller: So what happened?
Schweitzer: Well, for one thing, I think it was discouraging to a lot of dispensers that many consumer complaints were not mitigated by the routine use of probe-microphone testing.
Mueller: And you believe that's why probe-mic measures never became a routine method of hearing aid verification?
Schweitzer: A big part, yes, but recall that it was about the same time in the 1980s that we started seeing “programmable” hearing aids, led by 3M and Ensoniq. I probably made some crazy predictions about these two now-defunct hearing aid manufacturers too! It was a classic case of one reasonably important innovation getting promptly supplanted by another coming along on a different trajectory. As soon as hearing aids were more easily redesigned in the office by the dispenser, the ambiguity of a less clearly defined electroacoustic picture was taken to be a peculiar kind of virtue. A single performance “snapshot” of ear canal performance was thought to be less useful if the fitting was a more changeable pattern, re-adjusted on some presumed course of evolving improvements in the poorly termed “fine-tuning” process.
Mueller: So what does the future hold?
Schweitzer: I continue to believe in the value of probe-mic measurements, but, while many dispensers appear reluctant to stick probe tubes into their patients' ears, I'd at least like to imagine that they are listening more to consumer comments and making more adjustments according to user preferences. I wish we could teach it, but I certainly won't make any predictions about it.
Mueller: Yeah, let's stick with Niels Bohr this time!