An assessment of the validity of a subjective hearing aid verification procedure Whitney M. Vineyard, B. A. AuD Research Project April, 211 University of Memphis Project Advisor Robyn M. Cox, PhD.
Purpose Hearing aids are fit according to rules derived from evidence-based prescriptions. The best way to evaluate a hearing aid is using objective measurement, using real ear measures. However, many audiologists do not use real ear measurements to verify that the hearing aid is working according to fitting goals.
Research Questions 1. How well does a structured subjective test battery work to verify that a hearing aid is matching fitting goals? 2. How much time does it take to run a thorough subjective test battery? 3. What are the limitations of this kind of approach?
How fitting goals differ from NAL-NL1 NAL-NL1 prescription targets do not exactly align with fitting goals at MSHC, where we aim to: Match 55 db targets. Fall 5 db below 7 db targets. Verify MPO targets primarily through subjective verification. We use subjective verification procedures to verify that hearing aids are set to appropriate loudness levels.
Methods The test battery was performed using two different hearing aids on each subject. The test battery consists of three steps: 1. repeating speech sounds to assess soft speech audibility 2. rating average speech loudness and 3. rating loud sound comfort. The subject was asked to repeat sounds or rate loudness comfort. After completion of the test battery, the output of the hearing aid was compared to NAL-NL1 prescription targets which were generated by entering hearing thresholds into a Verifit system.
Participants Subjects were users of hearing aids with sensorineural hearing loss and no history of middle ear pathology or ear surgery. Subjects were recruited from the Memphis Speech and Hearing Center and Hearing Aid Research Laboratory. 2 Subjects 11 male, 9 female Average age: 76 years Average length of hearing aid use: 6.6 years. Average amount of wear per day: 11.2 hours.
Flow Charts In order to standardize the test battery, flow charts were developed. These charts clarified the adjustments that were to be made in the software based on the responses of the subject.
Step 1: subjective verification of soft speech audibility START: Administer 6 sound test. Did subject hear /a/,/i/, /m/ & /u/? Yes Did subject hear /sh/ correctly? Yes No No Is this > 3 rd repetition Yes Is this > 3 rd repetition Yes No No Turn up overall gain for soft sounds 5 db and readminister test. Turn up gain for soft sounds above 2 Hz 3 db and re-administer test. Turn up gain for soft sounds above 4 Hz 3 db and readminister test. No Is this > 3 rd repetiti on Yes No Did subject hear /s/ correctly? Yes Move on to STEP 2
Step 2: subjective verification of average speech loudness Administer test: ISTS track played at 65 db SPL to represent average speech Have the subject give a loudness category Refer to Table 1 for adjustment information based on level of experience Is the sound > the goal? No Is the sound < the goal? No Yes Yes Is this > 3 rd repetition Yes Is this > 3 rd repetition Yes No No Turn DOWN loud gain in all frequencies for average sound 3 db Turn UP loud gain in all frequencies for average sound 3 db Move on to STEP 3
Step 3a: subjective verification of loud sound comfortbroadband MPO Administer test: Broadband MPO Sound Have the subject give a loudness category Refer to Table 1 for adjustment information based on level of experience Is the sound > 6? No Is the sound < 6? No Yes Yes Is this > 3 rd repetition Yes Is this > 3 rd repetition Yes No No Turn DOWN overall MPO by 3 db Turn UP overall MPO by 3 db Move on to STEP 3b
Step 3b: subjective verification of loud sound comforthigh frequency MPO Administer test: High Frequency MPO Sound Have the subject give a loudness category Refer to Table 1 for adjustment information based on level of experience Is the sound > 6? No Is the sound < 6? No Yes Yes Is this > 3 rd repetition Yes Is this > 3 rd repetition Yes No No Turn DOWN MPO > 3kHz by 3 db Turn UP MPO > 3 khz by3 db Test complete
Starkey Inspire Software For Step 1, adjustments were made using the Soft Sounds Channel. For Step 2, adjustments were made using the Loud Sounds Channel. For Step 3, adjustments were made using the MPO arrows.
Siemens CONNEXX Software For Step 1, adjustments were made using the Soft sounds arrows. For Step 2, adjustments were made using the Speech arrows. After review, examiners realized that post-programming results were not in line with the goals of the verification battery. As a result, this protocol was restructured after 1 subjects. For Step 3, adjustments were made using the output limiting bar.
Siemens CONNEXX Software For Step 1, adjustments were made using the Master gain arrows. For Step 2, adjustments were made using the Loud sounds arrows. For Step 3, adjustments were made using the output limiting bar.
Audiometric Data, all test ears 1 2 3 4 5 6 7 8 9 1 11 12
Results 1. How well did the structured subjective test battery work to verify that a hearing aid is matching fitting goals? The accuracy of the fitting was evaluated by subtracting post-programming REAR from the individual s NAL-NL1 prescription targets across the frequency range.
db difference from target db difference from target Difference from target after verification, 55 db Starkey Siemens 25 5 1 2 4 25 5 1 2 4 15 15 1 1 5 5-1 -1-2 -2-3 -3
db difference from target db difference from target Difference from target after verification, 7 db Starkey 15 1 5-1 -2-3 -35 25 5 1 2 4 Siemens 15 1 5-1 -2-3 -35 25 5 1 2 4
db difference from target db difference from target Difference from target after verification, MPO Starkey 25 5 1 2 4 5-1 -2-3 -35 Siemens 25 5 1 2 4 5-1 -2-3 -35
Results 1. How well did the structured subjective test battery work to verify that a hearing aid is matching fitting goals? Question: Did the subjects move closer to fitting goals or further away?
Difference from Rx target (db SPL) Difference from target before vs. after, 15 55 db Starkey 25 5 1 2 4 1 5-1 -2-3
Difference from Rx target (db SPL) Difference from target before vs. after, 15 55 db Siemens 25 5 1 2 4 1 5-1 -2-3
Difference from Rx target (db SPL) Difference from target before vs. after, 2 15 1 5-1 -2-3 7 db Starkey 25 5 1 2 4
Difference from Rx target (db SPL) Difference from target before vs. after, 15 1 5-1 -2-3 -35-4 7 db Siemens 25 5 1 2 4
Difference from Rx target (db SPL) Difference from target before vs. after, MPO Starkey 5 25 5 1 2 4-1 -2-3 -35-4
Difference from Rx target (db SPL) Difference from target before vs. after, MPO Siemens 5 25 5 1 2 4-1 -2-3 -35
db difference from target db difference from target Difference from target before vs. after, 55 db Starkey First 1 Subjects Last 1 Subjects 25 5 1 2 4 25 5 1 2 4 15 15 1 1 5 5-1 -1-2 -2-3 -3
db difference from target db difference from target Difference from target before vs. after, 55 db Siemens First 1 Subjects Last 1 Subjects 25 5 1 2 4 25 5 1 2 4 2 15 1 5-1 -2-3 -35 15 1 5-1 -2-3
db difference from target db difference from target Difference from target before vs. after, 7 db Starkey First 1 Subjects Last 1 Subjects 25 5 1 2 4 25 5 1 2 4 15 15 1 1 5 5-1 -1-2 -2-3 -3
db difference from target db difference from target Difference from target before vs. after, MPO Siemens First 1 Subjects 25 5 1 2 4 Last 1 Subjects 25 5 1 2 4-1 -1-2 -2-3 -3-35 -35
Statistical Analysis T-tests and ANOVA tests were performed to evaluate the relationship between the preverification fitting and the post-verification fitting. Question: Were the adjustments that we made to the hearing aids statistically significant differences? Answer: Yes, for many frequencies a statistically significant change occurred. This means that the result was likely a product of our test battery, and unlikely due to chance.
Summary of Statistics Subjective Test Starkey Siemens Soft Speech Closer to target Closer to target Average Speech MPO Closer to db re: target fitting goal Further below target NO CHANGE Further below target For all frequencies where a significant change occurred, it brought the hearing aid output closer to fitting goals.
Summary of Results Results showed that for both hearing aids, soft speech moved closer to targets. Average speech results were mixed, with one hearing aid moving closer to db re: 7 db targets, and the other with no change. MPO moved further below targets (closer to fitting goals) for both hearing aids.
Results: 2. How much time does it take to run a thorough subjective test battery? For an average subject It took 8.5 minutes for the Starkey aid, and it took 7.5 minutes for the Siemens aid. 14 12 1 8 6 4 2 Siemens 8 Starkey 9 12 13 14 15 16 17 18 19 2 21
Results: 3. What are the limitations of this kind of approach? Results were dependant upon the motivation and cognitive ability of the subjects. Variability across the subject range increased after verification. It is important to really know what changes the software is making on the hearing aids! For example, the speech adjustment bar for one hearing aid software made changes across the entire dynamic range.
Conclusions Subjective verification has the ability to bring hearing aid wearers closer to fitting goals when real ear verification is not possible. Subjective verification worked better for one hearing aid than another. It takes about 8 minutes to perform the subjective test battery.