It was in 1979, at the International Ear Clinics' Symposium, that Earl Harford, PhD, first described a new technique to assess the performance of hearing aids by putting a microphone in the ear canal so that actual real-ear gain and output could be measured. Many believed that the “science” component of fitting hearing aids had finally arrived.
By the mid-1980s, leading audiologists were predicting that real-ear probe-microphone measures would be used by nearly everyone by the end of the decade. But that didn't happen. In fact, it didn't happen in the '90s or in the '00s either, despite Best Practice Guidelines, such as those of the American Academy of Audiology (AAA), stating: “Prescribed gain (output) from a validated prescriptive method should be verified using a probe-microphone approach that is referenced to ear canal SPL.” That's a pretty unambiguous statement.
Here at HJ, we've periodically sampled the use of real-ear probe-mic measures through dispenser surveys. For example, in 1995 we found that “routine” use of these measures was reported by 54% of audiologists (n=134) and 18% of hearing instrument specialists (HISs; n=108), with an overall use rate of 39%. In 1999, we examined use rates for both groups, but limited it to those who owned or had access to the equipment. Even then, only 42% reported routine use.
Our 2003 survey showed an overall use rate of 37% (n=558 audiologists, 49 HISs). And finally, in a 2005 survey, we again examined the popularity of these measures, this time just among audiologists. The overall use rate was 34%. It was slightly higher (∼40%) for recent graduates (either masters or AuDs) and for experienced audiologists who had obtained their AuD through distance learning.
For some reason, or probably for many related reasons, using real-ear probe-mic measures for verification of hearing aid performance has never become the prevailing practice. Our past surveys suggest that only about 1/3 of dispensers have been using this verification approach routinely, with no meaningful upward trend observed.
SIGNS OF CHANGE?
So, after 30 years, are things starting to change? Maybe. Some articles from the past year certainly have stirred up lively discussions on the topic:
* In a July 2009 article, Consumer Reports concluded that about 2/3 of hearing aids are not fitted correctly, that audiologists and HISs do not routinely conduct probe-mic measures to assure that they are fitted correctly, and that probe-mic testing is a “must have” procedure for every consumer purchasing hearing aids.1
* In a recent issue of Audiology Today, Catherine Palmer, PhD, suggested that the failure to use probe-mic measures in the fitting of hearing aids is unethical practice.2 She based this on items from AAA's Code of Ethics such as “Members shall maintain high standards of professional competence in rendering services” and “Members shall provide only services and products that are in the best interest of those served.”
* Last month, Sergei Kochkin and colleagues published an extensive article in Hearing Review on hearing aid outcomes, based on the data from MarkeTrak VIII.3 These data show that: (1) hearing aid satisfaction is related to the testing conducted at the time of the fitting, (2) more testing leads to more satisfaction, and (3) probe-microphone measures are one of the tests that affect these results.
ANOTHER PROBE-MIC SURVEY
Given this increased focus on hearing aid verification, we thought it was time to conduct another survey on this procedure. Several questions on the topic were included in the annual Hearing Journal/AudiologyOnline dispenser survey reported on in this month's Cover Story.4
The e-mail survey drew responses from 640 practitioners, including 416 audiologists (71%) and 147 (29%) HISs. For the probe-mic portion, which we'll be talking about here, we used data only from U.S. practitioners who dispense hearing aids and who answered all the questions on the survey. This resulted in a total of 420: 309 audiologists (74%) and 111 (26%) HISs.
For survey questions directly related to probe-mic test procedures, we included responses only from people who said they owned or had access to the equipment. Also, people were instructed to respond as the questions applied to their adult patients. Respondents who reported seeing only pediatric patients did not take the probe-mic procedures portion of the survey. The questions on probe-mic procedures drew 358 responses: 258 from audiologists (72%) and 100 from (28%) HISs.
Our first question was: “How often do you use probe-microphone measures to verify and adjust the gain/output of the hearing aid on the day of the fitting?” As in our previous surveys, respondents were offered five choices for describing their use of probe-mic measures: Seldom or Never, Sometimes, About Half the Time, Usually, or Always or Nearly Always. The overall distribution of responses for this question is shown in Figure 1. Observe that the results are quite similar to those from past surveys. Perhaps most notable is that the majority (52%) of respondents said they used this verification procedure no more than Sometimes.
Consistent with earlier HJ surveys, we defined “routine use” as using probe-mic measures more than half the time. Using this criterion, Figure 2 illustrates the distribution between audiologists and HISs, both for all respondents and separately for those who have access to the equipment. Note that there is no significant difference between the two groups. If we look only at the findings for audiologists, routine use is somewhat higher than we found in 2005 (45% versus 34%). This could be because of the greater percentage of AuD audiologists in the present sample—in the 2005 sample, AuDs had a use rate of about 40%, but were a smaller portion.
Naturally, probe-mic use is greater among respondents with access to the equipment than among all respondents, who include about 12% without access. But surprisingly, it's not that much greater. As indicated in Figure 2, nearly half of audiologists and HISs who have access to probemic equipment don't use it routinely.
INSERTION GAIN VERSUS EAR CANAL OUTPUT
Probe-mic equipment can be used in two different ways to verify hearing aid fittings. In one approach, real-ear insertion gain (REIG) is calculated, and these gain values are used to determine the “quality” of the fitting, based on some pre-determined frequency-specific gain requirements. This requires measuring the real-ear unaided response (REUR) (or using an average REUR), measuring the real-ear aided response (REAR), and then calculating the REIG by subtracting the REUR from the REAR.
The other verification procedure is to use only the REAR. In this case, you're not interested in the “gain” obtained per se, but rather the absolute SPL that is delivered to the eardrum. Since it has become common to use a speech-shaped noise signal or real speech (recorded or live) as the input signal, this REAR verification approach is often referred to as speech mapping. The ear canal SPL values that are obtained are then compared to some pre-established level of acceptability, usually based on audibility or prescribed targets.
Historically, the REIG has been used for hearing aid verification, primarily because early prescriptive methods (NAL-R, POGO, Berger, Libby, etc.) had targets only for REIG. But now that DSL and NAL-NL1 are the prescriptive methods of choice, REAR targets are available, and probe-mic manufacturers include REAR targets in their fitting software. We questioned if this has led to increased use of the REAR as the primary verification tool. For this analysis, we excluded data from respondents who said they conducted probe-mic testing “Seldom or Never.”
While some dispensers use both methods, when asked their “primary” method, 78% said the REAR and 22% the REIG. With REIG calculations, it is possible to use an “average” REUR, which is stored in the equipment's software. It has been debated whether the “average” or “measured” REUR more consistently provides the “best fitting,” and opinions on this topic are split about 50/50. However, among our survey respondents we saw a strong consensus: 91% report using the measured REUR. The responses from audiologists versus HISs were nearly identical for all questions in this category.
TYPE OF INPUT SIGNAL
In the early days of probe-mic measures, people often used a swept-tone as the input signal. For various good reasons, tonal signals mostly have been replaced by broadband inputs. The only exception is in measuring the real-ear saturation response (RESR), where a tonal signal is needed to drive the hearing aid to its MPO. As mentioned earlier, when the REAR is used for verification, it's common to use either recorded or live speech.
To examine how common these new inputs have become, we asked the respondents who had already said they primarily used the REAR for verification to identify the input signal they most often used. The choices offered them were: swept pure tone, broadband noise/speech-shaped noise, real-speech recorded (presented from the probe-mic equipment), or real-speech live (dispenser's voice or voice of family member).
We found that a noise signal was the most popular, although real speech, both recorded and live, was also commonly used. If the two real-speech categories are combined, then 59% of audiologists and 50% of HISs use speech as their primary input signal. (Note that this is only for respondents who use the REAR as their primary verification tool.) These data revealed a significant interaction between audiologists and HISs. Audiologists reported greater use of recorded speech (49% versus 26%), while HISs reported greater use of live speech (24% versus 10%).
This difference in input selection preference between the two dispenser groups prompted us to examine if there might be gender effects. Of the audiologists, 77% were female while the HIS sample was only 35% female. In two previous HJ surveys, one related to data logging and the other to open-canal fittings, we found significant interactions for gender. There was a significant gender effect for this question too (for the audiology group), as Figure 3 shows. Female audiologists used recorded speech more than their male counterparts (53% versus 29%) and live speech less (10% versus 21%). In fact, the findings for the male audiologists for these two categories were not significantly different from either gender of the HISs.
GOAL OF PROBE-MIC TESTING
In an earlier pilot survey, using an open-ended question, we identified four main reasons that people use probe-mic testing during the verification process:
* To determine if the average speech spectrum is audible and falls within the patient's residual dynamic range.
* To determine if gain is appropriate for the patient's hearing loss.
* To match gain and/or output to the hearing aid manufacturer's desired targets in the fitting software.
* To match gain and/or output to prescriptive fitting targets (e.g., NAL or DSL) displayed by the probe-mic equipment.
This year, we asked respondents to select which of these four they viewed as their primary fitting goal. As Figure 4 shows, the overall responses were fairly evenly distributed, although it appears that the most common use was to ensure audibility of the speech spectrum. There were significant differences between audiologists and HISs in two areas: Audiologists reported using probe-mic testing more to verify a generic prescriptive method (NAL and DSL), 27% to 10%, and were much less likely to use it to verify a manufacturer's software target (8% to 25%).
We were somewhat surprised by how few respondents used these measures to verify a fitting target such as NAL or DSL. This prompted us to go to a different question in the survey, which specifically asked respondents if they used either of these prescriptive methods. What we learned was interesting: 79% of audiologists and 87% of HISs reported using these prescriptive methods routinely. However, of these respondents, only 59% of audiologists and 39% of HISs said they routinely conduct probe-mic measures. That's a puzzling disconnect—like saying you used a blueprint to build a house, but not a tape measure. Of the respondents who said they used the prescriptive methods routinely and also reported conducting probe-mic testing routinely, only 37% of audiologists and 35% of HISs said that their primary fitting goal was to verify these targets.
A CHECK FOR VALIDITY
One need only look at the unexpected results of various elections to see that surveys have some problems with validity. While we were careful not to ask leading questions, to arrange answers in different orderings, and so on, we realize that the validity of any survey of this type can be questioned. Respondents may not read the questions carefully, or they may simply check any answer to finish the survey quickly. They also might sense what their answer should be, and lean in that direction. This survey was particularly susceptible to the later bias, as the Consumer Reports strong endorsement of probe-mic testing was mentioned earlier in the survey.
To account for this, we inserted a question for respondents who said they did probe-mic testing with adults asking how often they performed different probe-mic procedures. We included common procedures such as the REUR, REAR, RESR, but we also asked about the “Binaural Summation Index”—a non-existent test. We chose this name because we assumed that anyone who really did do probe-mic testing routinely and read the question would know that (a) probe-mic testing is not conducted binaurally (bilaterally), (b) binaural summation is not something you can measure in the ear canal, and (c) binaural summation is not measured as an index.
Of the respondents who reported conducting probe-mic testing routinely, 21% of the audiologists and 28% of the HISs said they did this test at least sometimes. When we asked these respondents on how many of their patients they used this imaginary test, the audiologists gave an average response of 78% and the HISs 79%.
On this question, we also found what appeared to be an interesting “educational” effect. That is, 25% of master's audiologists reported doing the test, but only 12% of the AuDs. In case you're curious, we did have nine PhD audiologists complete the survey—four of them said they did the test! So, our survey isn't perfect, but we doubt that most are.
SUMMARY OF FINDINGS
* The routine use of probe-mic measures for hearing aid verification is around 40%, not much different from survey data over the past 15 years. There does appear to be a slight increase for audiologists since our 2005 survey.
* Having the equipment available does not ensure that people will use it. Nearly 50% of those who have the equipment do not use it routinely.
* The REAR is used three times as often as the REIG as the primary means of verification.
* When using the REAR, most dispensers use either real or live speech, although many also use a shaped-noise signal.
* The primary use of probe-mic measures is to assess audibility of the speech spectrum.
* Most practitioners report using a validated prescriptive method, yet no more than half of these people routinely use probe-mic measures to verify it.
* In general, the findings for audiologists and HISs did not differ much, except for the use of manufacturer's software targets vs. prescriptive algorithms and the use of live vs. recorded speech.
* In general, the findings were similar for men and women, except among audiologists in their use of live vs. recorded speech in doing REAR testing.