Getting Aided Loudness Right: A Three Step Approach



Author: Brian Taylor, Au.D. & H. Gustav Mueller, Ph.D.

Much of the initial acceptance and long-term success of a hearing aid fitting rides on the audiologist’s ability to get the maximum power output (MPO) of the hearing aids set correctly. While both wide dynamic range compression (WDRC) and AGCo easily can be adjusted with today’s instruments, the most recent MarkeTrak VIII survey reveals that a large percent of hearing aid users remain dissatisfied with their aided perceptions of loud inputs (Kochkin, 2010). We have to assume, that in many of these cases, this is because the maximum output of the hearing aids was not adjusted correctly. It is important, however, to separate the “annoyance” of loud sounds from “uncomfortable loudness.” (which we’ll discuss later). In this brief review, we’ll describe a three step approach that easily can be clinically implemented to assure that MPO setting are appropriate for your adult patients.

Step 1: MPO Adjustment Before the fitting
To ensure that loud sounds are not uncomfortably loud, the MPO needs to be set just below the measured loudness discomfort level (LDL). We say “just below” because setting the MPO too low can result in poor sound quality and may even compromise speech understanding in background noise [e.g., peaks of speech could be unnecessary limited, reducing the overall signal-to-noise ratio; (SNR)]. While the output also can be limited with the WDRC, normally we use the AGCo for this purpose. Getting the MPO right, therefore, means the audiologist must carefully set the kneepoints across multiple channels of AGCo.

As described in Best Practice Guidelines, and reviewed by Mueller (2011a), the first step in “getting the output right” is to assess unaided LDLs at one or two key frequencies, and then convert these HL values to 2-cc coupler [by adding the reference-equivalent sound pressure level (RETSPL)] for setting AGCo kneepoints. For example, if the patient’s LDL at 3000 Hz is 108 HL, the 3000 Hz RETSPL of 2.5 dB would be added (for 2-cc conversion) and a ~110 dB SPL value would result. It is then possible to go to the fitting software, and select this value for the MPO for 3000 Hz. This procedure can be repeated for desired key frequencies.

These HL LDL values also are then later converted to earcanal SPL [by adding the Real Ear to Dial Difference (REDD)], and can be used as the upper target level when RESRs measures are conducted. With multichannel AGCo, the RESR can be shaped to maximize headroom across frequencies without exceeding comfort levels.

Step 2: Aided MPO Verification
While these pre-fitting measures will very often lead to getting the AGCo kneepoints correct, we believe that there also is a need to conduct aided behavioral measures of loudness perception. There are several reasons why this might be appropriate (see Mueller, 2011b for review):
  • The unaided LDLs may not have been valid.
  • Individual differences in monaural summation (e.g. the difference between a tone and a broadband signal for an given patient).
  • The effects of bilateral summation (the result of differences in aided loudness perception when listening through two hearing aids compared to the loudness perception in the unilateral earphone unaided condition).
  • Hearing aid channel summation (the additive effect of output when a broadband signal is processed across several hearing aid channels).
When aided loudness verification is conducted we recommend using both speech signals and “obnoxious noises”. This agrees with the 1994 IHAFF protocol. This protocol recommends the use of the 7-point loudness anchors from the Cox Contour Test (Cox, 1995) to assess the perception of aided loudness in the test booth. (See pp. 346-347 of Taylor & Mueller, 2011 for an example of how to conduct this loudness verification testing).

Most all manufacturers have obnoxious noises as part of their fitting software, and probe-mic equipment also has such signals. If conducted using the fitting software, you will need a handy sound level meter to assure that the output is at an appropriate level (something around 85 dBA seems to work pretty well). This is not necessary when using probe-mic equipment, as the reference microphone will assure the signal is calibrated to a given level.

A somewhat more unique method for measuring aided loudness is the the HARL Obnoxious Noise Kit or HONK (Cox, 2009). The HONK is an obnoxious noise kit consisting of one 12-ounce metal coffee can with nuts and bolts inside, two glass jars with 24 marbles. Using the 7-point loudness anchors, the audiologist can have the patient rate aided loudness perception and make changes to the AGCo as needed to assure a “Loud, But Okay” rating.

Step 3: Real World Questionnaires
While the use of unaided LDLs for the initial setting, and using aided loudness testing as part of the verification process are good laboratory measures, they do not capture loudness perception in truly real-world listening conditions. It is still possible that your patients might encounter noises in the real world that are uncomfortably loud, or their overall judgments of what is too loud might change once they leave the sterile environment of the clinic. Additionally, even though these behavioral measures evaluate aided loudness they do not assess the individual’s satisfaction with their perception of loudness with hearing aids across several listening situations. Questionnaires, however, can be used to assess real-world perceptions of aided loudness.

A PAL for your clinic
The Profile of Aided Loudness (PAL) was introduced by Mueller and Palmer in 1998, with administration and scoring details provided the following year (Palmer et al, 1999). Additional data for test interpretation (loudness rating distributions for test items for normal hearing individuals) was provided in 2000 by Palmer and Mueller.

Since some of you may not have read the articles on the PAL in the late ‘90s, a little background on the development is in order. The selection of the 12 items on the final version of the PAL began with a list of 95 “sounds” that new hearing aid users had reported hearing during their first few weeks of hearing aid use. Normal hearing subjects between the ages of 20 and 65 rated the loudness of each sound based on the 7-point loudness scale of the Cox Contour Test.

The 12-items of the PAL include four items relating to soft sounds (target is a #2 rating), four items for average (target is a #4 rating) and four items for loud (target is a #6 rating). Because of the rigid criteria used for item selection, it’s possible that your patient may have a poor memory of the loudness of some of the items (e.g., a requirement for high test-retest reliability).

In addition to the loudness categories, there also are satisfaction ratings for each item ranging from #5 (best) to #1 (worse). We know that even when hearing aids setting appear to correct, the patient may still find certain sounds “annoying” and not be satisfied (Palmer et al, 2006). There are not norms per se for satisfaction, although when the PAL is used as an outcome measure, any satisfaction rating below a #3 (“okay”) needs to be discussed with the patient. The highest satisfaction rating of #5 is not necessary good (for the patient). For example, a patient could rate “Your Own Breathing” as a #0 (cannot hear; desired rating is #2) on the loudness scale, yet rate this as #5 (Just Right) for satisfaction.

The ESQ, a cousin of the PAL
In several studies of a hearing aid signal processing scheme termed ADRO (adaptive dynamic range optimization), Blamey and Martin (2009) used a modification of the PAL inventory to obtain loudness ratings and satisfaction. Their self-assessment scale is the Environmental Sounds Questionnaire (ESQ). The ESQ utilizes the same loudness and satisfaction scaling as the PAL, and includes nine of the 12 PAL items. Nine additional items were added for an ESQ total of 18. Interestingly, several of the nine new items of the ESQ were included in the original PAL research, but were not included in the final version as they did not meet the selection criteria. The items from the PAL that were deleted were:
  • The dryer running
  • The religious leader during a sermon
  • Listening to a marching band
The nine items added to the ESQ were:
  • Traveling in a car with the windows closed
  • Traffic noise when standing on the curb of a busy road
  • Washing machine
  • Running water, such as a toilet or a shower
  • A motorbike passing by
  • Vacuum cleaner
  • Telephone ringing close by
  • Refrigerator motor
  • Birds twittering
Administration and Scoring
The PAL is a “paper and pencil” administered scale, and can be conducted in the waiting room prior to the clinic visit, or it could be mailed to the patient. The PAL may also be given to the patient to take home and complete between appointments. Some audiologists administer it “knee-to-knee” by reading the questions to the patient. The 12 items from the three different loudness categories are randomized on the form (i.e., the four items with a normative rating of #2 do not follow each other).

The instructions are as follows:

Please rate the following items by both the level of loudness of the sound and by the appropriateness of that loudness level. For example, you might rate a particular sound a “Very Soft (#1). If Very Soft is your preferred level for this sound, then you would rate your loudness satisfaction as “Just Right.” If on the other hand, you think the sound should be louder than “Very Soft”, then your loudness satisfaction rating might be #2: “Not Too Good” or even #1: “Not Good At All.” The loudness satisfaction rating is not related to how pleasing the sound is to you, but rather, the appropriateness of the loudness.

Sample Item: The hum of a refrigerator motor:

Loudness Rating Satisfaction Rating
0. Do Not Hear 5. Just Right
1. Very Soft 4. Pretty Good
2. Soft 3. Okay
3. Comfortable, But Slightly Soft 2. Not Too Good
4. Comfortable 1. Not Good At All
5. Comfortable, But Slightly Loud
6. Loud, But Okay
7. Uncomfortably Loud


In the example above, the patient rated the loudness level of a refrigerator motor running as “Comfortable, But Slightly Soft,” and rated his loudness satisfaction for this sound as “Just Right.” This satisfaction rating indicates that the patient believes that it is appropriate for a refrigerator motor to sound “Comfortable, But Slightly Soft.”

It is also important to inform the patient that it is okay to skip an item if it’s simply not something they recall hearing, and don’t have a loudness perception.

Figure 1. Normative Data for PAL


Once the patient has completed the inventory, scoring is quite straightforward. Using the form shown in the Appendix, enter the patient’s loudness and satisfaction ratings for each of the 12 items, and then calculate an average of the four.

In some cases it might be useful to go into more detail regarding the patient’s judgments. The distribution table shown in Figure 1 can be helpful. These are the data from the original normal hearing subjects used in the development of the PAL. The bars displayed for each item represent the percent of subjects that gave the item this rating (e.g., about 75% of individuals consider a marching band to be “Loud, But Okay”). The item distributions in Figure 1 are especially useful for counseling the patient when the PAL is used as an outcome measure, but also is helpful during pre-fitting. For scoring, we usually just put an “X” for the patient’s rating for a given item. It easily can then be visualized what type of amplification needs are present. For example, note that this patient rated the loudness of a “Religious leader during the sermon” as #2 (Soft). The majority of people with normal hearing, however, rate this item#4 or #5, and only 2-3% rate it the same as this patient.

The APHAB Aversiveness Scale
The Aversiveness subscale of the Abbreviated Profile of Hearing Aid Benefit is another questionnaire or self-report that can be used to evaluate loudness perception in real-world listening conditions. Here are the questions on Form B of the Aversiveness subscale of the APHAB.

Aversiveness (AV Scale)

1. The sounds of a fire engine close by is so loud that I need to cover my ears.

8. The sounds of running water, such as a toilet or shower, are uncomfortably loud.

11. The sounds of construction work are uncomfortably loud.

14. Traffic noises are too loud.

18. Unexpected sounds, like a smoke detector or alarm bell are uncomfortable.

22. The sound of screeching tires is uncomfortably loud. Recall that the APHAB is scored on a 7-point Lickert scale. The APHAB, as well as the normative data for it, can be obtained at http://www.memphis.edu/ausp/harl/aphab.htm.

You may want to consider using both the PAL and APHAB as outcome measures. If you conduct both the PAL and the APHAB you would expect to see agreement between the PAL ratings for the four “Loud, But Okay” questions and the APHAB subscale score for aversiveness.

An important thing to remember regarding APHAB aversiveness scores is that using hearing aids usually will make things worse, especially for new hearing aid users. What you might see, and what might be your fitting goal, is that the aided aversiveness scores are fairly similar to those of people with normal hearing.
Clinical Use of the PAL
During the pre-fitting assessment, using the data in Figure 1, the patient can be counseled regarding how they differ from people with normal hearing. The PAL findings might also serve as a warning of potential loudness problems down the road. For example, if all the items that typically are rated #6, are rated #7, it may be important to assure that little or no hearing aid gain is provided for inputs of this intensity range. One of the most important reasons for using the PAL as a pre-test is to establish a baseline for which aided PAL responses can be compared. Once your patients start using hearing aids, it’s a bit risky to then have them go back and rate how they heard when they were unaided. They may find that the sounds they once thought were “comfortable”, actually were really “soft.”

Some time after the fitting of hearing aids (30 or so days post fitting), the PAL can be re-administered in the aided condition. On page 2 of the PAL form, there is an area to compare unaided to aided performance. You can find the PAL at the end of this article. We encourage you to make copies and use it during your next fitting as we describe here. When individual scores deviate from the target scores for soft, medium or loud sounds, or satisfaction ratings are less than 3, audiologists need to adjust WDRC or the AGCo kneepoints, and counsel the patient accordingly. As reviewed by Mueller and Palmer (1998) there are four general categories for outcomes for the PAL:

Outcome #1: Goal of normal aided loudness perception met, patient is satisfied. Give yourself a pat on the back. Life is good!

Outcome #2: Goal of normal aided loudness perception met, patient is not satisfied A common finding for a new hearing aid user. Depending on your belief of “who knows best,” this could be either a hearing aid adjustment trigger, or it might simply mean more counseling is needed. Assuming the patient is dissatisfied because things are too loud, you might want to attempt to discern if these sounds are truly uncomfortable, or are they annoying? Follow-up aided loudness testing could be conducted. Also, consider that Mueller and Powers (2001) report changes in the PAL following acclimatization.

Outcome #3: Goal of normal aided loudness perception not met, patient is satisfied. An interesting scenario. Do you really want to change programming for a happy patient? Your decision will probably depend on what loudness levels are rated inappropriately. See Mueller (1991) for clinical guidelines.

Outcome #4: Goal of normal loudness perception not met, patient is not satisfied. Time to start over. Repeat hearing aid programming, probe-mic verification and aided loudness tested. Repeat the PAL in a few weeks.

In Closing
Getting the output right is an important component of the overall hearing aid fitting protocol. We believe that it starts with LDLs obtained with earphones. These values can then be used for the intital MPO settings. At the time that the hearing aids are fitted, RESR measures can be used for verification of loud sounds. During the subsequent follow-up period, the PAL as well as other behavioral measures of loudness can be used to validate aided loudness perception.

Restoring loudness perceptions to acceptable levels in the real world is also an important part of the overall fitting. The PAL can be used to validate that acceptable aided loudness perception has been restored and that the patient is satisfied.

Overall, it’s a fairly simple three-step process, which we believe will lead to greater patient satisfaction with hearing aids.    
References
Blamey, P. & Martin, L. (2009) Loudness and satisfaction ratings for hearing aid users. Journal of the American Academy of Audiology. 20, 2, 272-282.

Cox, RM (2009). Verification and what to do until your probe-mic system arrives. The Hearing Journal. 62, 10, 10-14.

Cox, RM. (1995). Using loudness data for hearing aid selection: The IHAFF approach. The Hearing Journal. 48, 2, 39-44.

Kochkin, S., Beck, DL., Christensen, LA., Compton-Conley, C., Fligor, BJ., Kricos,P., McSpaden, JB., Mueller, HG., Nilsson, MJ., Northern, JL., Powers, TA., Sweetow, Taylor, B. & Turner, RG.,(2010): MarkeTrak VIII: The Impact of the Hearing Healthcare Professional on Hearing Aid User Success, The Hearing Review. 17, 4,12-34.

Mueller, HG (2011a). How loud is too loud? Using loudness discomfort level measures for hearing aid fitting and verification, Part 1. Audiology Online, published 6/20/2011. Downloaded 1/17/2012.

Mueller, HG (2011b). How loud is too loud? Using loudness discomfort level measures for hearing aid fitting and verification, Part 2. Audiology Online, published 6/20/2011. Downloaded 1/17/2012.

Mueller, H.G. & Powers, T. (2001). Consideration of auditory acclimatization in the prescriptive fitting of hearing aids. Seminars in Hearing, 22, 2, 103-124.

Mueller HG. (1999). Just make it audible, comfortable and loud but okay. The Hearing Journal, 52, 1,10-17.

Mueller, HG & Palmer, CV (1998). The profile of aided loudness: a new “PAL” for ’98. The Hearing Journal. 51, 1, 10-19.

Palmer, C., Mueller, HG, & Moriarty, M. (1999). Profile of aided loudness: a validation procedure. The Hearing Journal, 52, 6, 34-42.

Palmer, C. V., & Mueller, H. G. (2000). Hearing aid selection and assessment. In: Alpiner, J. G., & McCarthy, P. A. (eds). Rehabilitative audiology, children and adults, Baltimore, Lippincott Williams and Wilkins, pp 332-376.

Palmer C, Bentler R, Mueller HG. (2006). Amplification with digitial noise reduction and the perception of annoying and aversive sounds. Trends in Amplification, 10, 2, 95-104

Taylor, B. & Mueller, HG (2011). Fitting and dispensing hearing aids. San Diego, CA, Plural Publishing.

Brian Taylor, Au.D. is director of practice development and clinical affairs for Unitron and editor of Audiology Practices. Dr. H. Gustav Mueller, Ph.D. is professor in the Department of Hearing and Speech at Vanderbilt University, Senior Audiology Consultant at Siemens Hearing, and Contributing Editor at AudiogyOnline.