Tips on making the most of new hearing aids during the trial

Sound recover should be verified.

If you found the domes uncomfortable then they should be remade, and if you feel like they are making you hear worse that should be identifiable–did they do a direct comparison with REM to see where the differences were?

But 50% to 67% is not actually a statistically significant difference on a WRS test, which is one of the reasons it’s a poor test. I trust that there was a real difference, but only because you say so and your reported experience is reliable.

Well, how does REM work with Sound Recover? Does it disable SR? The REM matches the target. If i have tones at 3000 and 4000 but 2 dead regions (DR) in between 3k and 4k, target still thinks i can hear there. Say my loss is 80 at 3000 and 90 at 4000. Would Target think my loss is 85 at 3500, where it may be 120? So target is 85, REM is 85, but im not hearing it.

That’s why i think aided word recognition is important. It could verify all is well, or it could show the person is not getting expected results.

You think that aided wrs would be more meaningful than your own self-report? Humans are phenomenal pattern detectors. If you say that you are hearing better one way, you are correct and the job of the audiologist is to trust you, and if they are curious to found out why. I’m saying that aided wrs doesn’t add anything further to your report, given it’s wide variability. You can detect smaller changes than it can detect. If a statistical non-difference in aided wrs helped your clinician believe you I suppose that’s good, but she should have believed you in the first place. That’s not to say there’s no place for aided wrs, but this isn’t a situation where it is clinically useful.

Verification of /s/ audibility confirms that sound recover is doing it’s job. Looking at your audiogram, the lowering would likely need to be below 2khz in the left and even stronger in the right to achieve /s/ audibility, so dead regions above that are neither here nor there.

Thanks for your insight, Neville! I’m currently at 1300 on Sound Recover for both ears and that seems to be a good spot. I still get very good results on one-to-one conversations in quiet environments so i think i will delay a CI for another year or so.

I’m still thinking it’s spurious: 24% is getting into good guesses plus a bit of transcranial effect.

I think it’s easy to forget the geometric/log progression of asymmetric sensitivity because the loss is plotted in a more linear fashion. The ears have approximately a 64,000 times difference in my book.

We already know that using REM targets, I get unacceptable distortion of vowel sounds. So we have to always be under REM targets, and my audi has given up on doing REM. But I’ve been trialing different hearing aids, and get the feeling that I’m having more difficulty distinguishing words with my most recent ones. I’d like to have an aided WR test to be able to compare results with one type of aid vs another, in alll cases after they have been tweaked to get the best result we think we can get.

@Jody said:

I’d like to have an aided WR test to be able to compare results with one type of aid vs another, in alll cases after they have been tweaked to get the best result we think we can get.

Okay. Just remember that on a full 50-word NU6 recorded list, if your first score is 66% and your second score is 80%, that’s not actually a statistically significant difference due to test-retest variability within the WRS test, and I believe that’s under inserts–variability on the aided test would probably be higher, and variability with a 25-word list will also be higher.

You see what I’m saying? You are better at detecting small but meaningful differences than the test is. So if your word recognition with no hearing aids is 0% and with hearing aid brand A is 72% and with brand B is 84%, this is a clinically meaningless difference and may not reflect differences in your real-word performance with the devices. So what do you DO with that information?

You can barely use it as an A:B comparison- even with recordings. Manually reading word lists should just be binned: full stop.

Testing with inserts and cans just removes the different pinna emphasis of different words, while Free Field testing (aside from the obvious stereo advantages) is practically impossible to replicate. Your client leans forward a foot, or turns 15 degrees to one side or another and your results are flawed. Add in noise field testing and you add another variable parameter too.
That’s before you consider client fatigue, hunger and dehydration plus the influences of prescription medication over a day.

Years ago, I worked for Knowles, with a device called KEMAR in an anechoic chamber with 1m deep cones in the walls, ceiling and floor - you basically worked in a 9foot cage in the middle. You wouldn’t believe the kind of minor changes in set-up that would give testing variability of a few dB. There’s no way your average test-room set up can notify single digit dB changes; especially using words, not tone sweeps.

@Um_bongo

I’ve had HA’s for 20+ years. All of my audiologists have used manually read word lists. My new hearing instrument specialist reads the words. In fact his list is shorter…

My issue has become word recognition. Before the setup was so bad I couldn’t hear in quiet, let alone noise.

I really appreciate your posts. And please keep itup!

Eleven audiologist.
Ten read the words.
Nice and loud.
Last one used automated words.
Picked up on the left ear distortion.
Didn’t realize what it was until I told her.
Just new it was worst than my right.
My audiogram says there about equal.