Why Not Word Recognition Tests After Fitting With Hearing Aids


I have had many hearing tests. Of course with tones and tones in noise, and sometimes word recognition with and without noise. What I never have had is those kind of tests AFTER I had been fitted with the aids. Instead the audiologist lets us go home and wait until we complain. Why do they not test us with the aids in place and make changes while we are there.


I’ve said the same thing. You want to sell me hearing aids? Show me the difference they make in word recognition or even pure tone test.


Another thought on the subject of testing after the aids are fitted: maybe a little more time spent by the audiologist in this way would diminish the high return rates that have the effect of raising the cost of hearing aids.


Although this sounds great in theory, I think there are a couple problems. 1) The test is not very precise. Results that vary by 10-20% are not really considered significant. 2) The expected benefit is small, perhaps too small to even reliably detect, expecially in light of number 1.


You may be right, but what would be the difference between someone saying they didn’t understand certain words (at the time of the fitting), and having them come back later saying the same thing? You have the additional lag of having to make an appointment.


Odds are that you’re still not going to understand some words. I think that applies to normal hearing too. I think the audiologist would want to know 1) if things are improved and 2)what situations, sound or words you consistently have problems with. People really want ears to be like eyes. Take a test. Correct vision and then pass test with 20/20 vision. Unfortunately that is not how hearing works. I do think many audiologists are doing informal testing with you just in their conversations with you.


Another possibility that could be a little more precise would be to have an audiologist play a section of a DVD (possibly repeated with different levels of noise injected). Of course with the newly fitted hearing aids. The audiologist would be able to control this environment in terms of volume and noise. The listener would then be required to press a button when he didn’t understand a word. This would inject some precision into the test and could allow the audiologist to make changes in the settings. It seems it could also be useful to use these responses to build a database of the hearing discrepancies as well as the setting which corrected those deficiencies.


Let me add another factor here. I am not saying this in terms of bragging, but more to add some credence to some of my ideas: I am a retired engineer with four patents(none is the field of hearing). This record should show that I have a history of thinking through problems/issues and coming up with a fix.


The tester has to know what you think you heard, which may not be what was said. Like those cartoons where a bunch of old people are chatting and comically misunderstanding each other.


My Audi occasionally tests my hearing with my hearing aids in and compares the word recognition scores to the unaided scores. This comparison is especially useful when my hearing has declined or I feel I need adjustments. It is taken in the soundproof booth, so the scores are better than in real life, where there is noise, accents, people mumbling, poor lighting, etc --all the things that hinder us further in understanding speech.


That should be relatively easy with a button press. It would just pause the recording momentarily and record what you thought you heard.


No, we are not comparing it to eye tests. But, the word recognition scores are quantified, as a percent. If you are saying the new hearing aids will help my word recognition, then quantify it. If my unaided score at 55db is 50%, and aided scores are 85% with aid A and 95% with aid B, you just sold me aid B.


I wish you the best of luck in your endeavor to develop tests that will make hearing aid fitting easier and more effective. The closest things I know of that are similar to what you mentioned are speech in noise tests, like Quick SINT. As mentioned, word recognition tests just give general indications. For people with a high frequency loss, a test of being able to differentiate plurals might show more differentiation.


@MDB: Aided hearing speech testing is called “validation” (of benefit); and it differs from real ear “verification” (to arbitrary prescriptive recipes), of which I believe has limited benefit.

Oh, And By The Way, any hearing professional who does not validate reverse slope (low frequency) hearing loss fittings is doing their patients a dis-service.


I don’t plan to try to develop that kind of test personally. I just wondered why that kind of test wasn’t used. One person who responded said his audiologist did that. I think some audiologists follow this forum, and I hoped maybe they would respond. As I said, the testing that I have had done at Costco didn’t do word recognition tests - even before the aids were fitted. Other audiologists, especially those associated with ENTs, did word recognition testing.


Curious how the demonstration of benefit is done. Is it done via word recognition scores? If so, at what kind of volume.


Are you saying that Costco didn’t do Word Recognition Testing as part of the hearing test. Every hearing exam I’ve had did Word Recognition Testing. I’ve had it done once with hearing aids at an academic cener. Really didn’t seem to be of much use.

Sorry for my misunderstanding. It really sounded to me that you thought you could come up with a fix.


I hope I’m not getting early altzheimers here, but to my recollection in the 8 years that I have had my hearing tested at Costco, I can’t ever recall a word recognition test there.


On my Costco paperwork, the audiogram takes up the top half of the page. SRT and Word recognition scores are on the bottom part of the page, along with space for other results.


Because someone above asked for an audiologist to weigh in…

For word recognition scores specifically: It’s not a very sensitive test. A 50 word list can require upwards of 20% change to be significantly different. It is also not a test that is really intended to give that much information about how well you are hearing at a conversational level, although practitioners certainly use it that way, and use it as a counselling tool. Certain patterns in word recognition scores are evidence of retrocochlear lesions and I am primarily looking for these red flags when I run it. WRS can also give a vague idea of maximum performance when presented at high levels but, excepting certain hearing loss configurations, the addition of hearing aids wouldn’t affect this score much.

Other unaided vs aided testing: Practitioners can and do do this. Aided threshold testing used to be the norm before real-ear verification became a widespread option. Pre- and post- speech-in-noise measures can be taken, although real-world generalizability is weak. As for comparing one hearing aid to another, you run into variability problems again. If you got 85% with hearing aid A and 95% with hearing aid B, that’s probably due to test-retest variability and learning effects. If you get 50% with hearing aid A and 95% with hearing aid B, well you can report that to me anyway. Your own experience is a more sensitive test.

(And to express a probably unpopular opinion: While the first-fits of two different hearing aid manufacturers can be dramatically different, if they are matched for output they become very similar, particularly in the controlled enviroment of the clinic. I would also suggest that OFTEN when someone likes hearing aid B more than hearing aid A, part of that is because the practitioner learned something about the patient’s listening preferences between trialing A and B and adjusted hearing aid B differently.)