By Ted Venema
Ted Venema
In the beginning was Functional Gain. Real-Ear measures did not exist. All hearing aids were Linear.
Sam Lybarger, whose inventions revolutionized hearing care in the 20th century, stood a Texas yard from the listener who wore the hearing aid, spoke in a normal conversational voice, and asked what sounded comfortably loud. He found the listener wanted gain that was close to about ½ of the hearing loss. Thus the “½ gain rule” was born.
FUNCTIONAL GAIN MEASURES BECOME THE NORM
Aided thresholds in a sound field with a hearing aid set at a comfortable volume control setting were compared to unaided thresholds measured with headphones. A successful fitting was signified by little letter A’s written across the audiogram, showing a lift of thresholds about halfway up toward the 0 dB HL line. The idea was that average speech inputs, plus the hearing aid gain, would give an output that fell within the client’s dynamic range (Figure 1).
Figure 1. The A’s stand for “aided thresholds.” The goal was to raise the thresholds by roughly ½ (and less than ½ in the lows to reduce the upward spread of masking). This way, speech inputs, plus the ½ gain, would produce aided speech outputs that sat within the dynamic range and did not exceed LDLs.
Fitting methods evolved from various different philosophies (Berger, POGO, Libby, NAL), and so exactly where you’d want the little letter A’s to appear on the audiogram would differ slightly from method to method. All fitting methods, however, had as their spinal cord, the ½ gain rule.
Then, in the late 1980s, came Real Ear. Fitting methods did not change much, but Insertion Gain became the order of the day. It was faster, and it yielded objective results. The whole idea was to compare the Real-Ear Unaided Response (REUR) with the Real-Ear Aided Response (REAR), with the difference between them being the Real-Ear Insertion Gain (REIG). If this matched the Fitting Method Target that was instantly calculated from the audiogram you had entered, you were good to go (Figure 2)!
Figure 2. Note that the #’s for the Real-Ear targets (asterisks) here are identical to those for the A’s on the Fig. 1 audiogram. Real-Ear Insertion Gain measures, however, are non-behavioral and faster. The audiogram is nowhere to be seen.
But just try counseling a client from this perspective: “Well, you see, this line is what we’re supposed to hit and this little lighter line is right near it, so your hearing aid is doing what it’s supposed to do.” The main problem was that the audiogram was not visually part of the picture, so aided speech outputs had to be imagined.
Interesting too, was that REUR wasn’t incorporated at all in the unaided testing under headphones, but, oh well. Non-behavioral Real-Ear measures were a whole lot faster than testing someone’s thresholds twice! Another good thing about good old Insertion Gain was that if someone came in saying the new hearing aid just didn’t sound like the old one, you could do a quick Real-Ear measure on the old one, and then make the new hearing aid do the same thing. That was much better than relying on, “How does that sound?”
IN SITU OUTPUT ENTERS THE PICTURE
Richard Seewald really is the father of newer Real-Ear measures. His DSL (Desired Sensation Level) Fitting Method arose in the early 1980s, and with it, the SPL-o-Gram. Insertion Gain and REUR were unceremoniously tossed onto the garbage heap of audiologic history.
Now the whole focus was on In Situ Output, also known as REAR (Real-Ear Aided Response). Trouble was, only Seewald and his followers used the SPL-o-Gram and DSL. Most clinicians, including myself, plodded on with Insertion Gain Real-Ear measures.
The National Acoustic Laboratories’ NAL-NL1 gradually began to follow the lead of the DSL’s SPL-o-Gram. One could initially see NAL’s simultaneous usage of both Insertion Gain and In Situ Output. This was followed in about a year, however, by a rather quick dumping of Insertion Gain. For DSL then, imitation could be considered the sincerest form of flattery.
Figure 3. The SPL-o-Gram shows the audiogram, and also the targets in terms of In Situ Output. Note the three targets and how each is generally placed within the client’s dynamic range. All Fitting Methods seek to accomplish roughly these same objectives.
Let’s look at the SPL-o-Gram (Figure 3). Everything is plotted according to Output, and in terms of dB SPL, so now hearing loss and hearing aids are speaking the same language. “More” on the graph now goes up, like every other graph in the world (except the Oddiogram). Normal hearing thresholds are placed on the bottom and Loudness Discomfort Levels (LDLs) are placed on the top. The patient’s hearing loss is placed part way up on the graph, thus showing a reduced dynamic range.
Since Compression hearing aids give different gains and output for different input levels, three targets are shown for: soft, medium, and loud inputs. The idea is to aid the listener so that soft speech input sounds soft, average speech input sounds average, and loud speech input sounds loud. Now there’s an improvement for counseling! One can compare unaided to aided speech. Clients can readily see what parts of speech were inaudible without hearing aids, and what parts have now become audible when aided. As we say in Canada, “Neat, eh?”
The rub is that all Fitting Methods were actually trying to accomplish what the SPL-o-Gram shows! In the past, however, we didn’t have the equipment to show it. Functional Gain compared aided to unaided thresholds, and Insertion Gain compared REAR to REUR. Gain was the order of the day—but not any more! Visualizing (1) normal hearing, the client’s audiogram, and the reduced dynamic range; and (2) unaided and aided speech outputs all on one graph really was a giant leap forward for audiology
However, there’s another twist to this story. Fitting Methods (e.g., NAL-NL2 and DSL 5 for adults) have become so similar that if you don’t compare them carefully, you may not notice the differences.
Check out the target comparisons for yourselves. If you place soft input speech so that when aided the output speech surrounds the thresholds, you’ll find that the patient can barely hear it. That’s normal; neither can you or I. Average speech inputs should be aided so its outputs sit in the dynamic range about 1/3 above the thresholds. Loud speech inputs should be aided so they sound loud, but remain below LDLs. Isn’t that what Lybarger would have wanted? Isn’t that what all Fitting Methods are trying to do in the first place?
“THE BOTTOM LINE FOR EVERYONE”
For adults at least, it looks as if Fitting Methods are actually becoming obsolete. To borrow a phrase from the songwriter Paul Simon, they are becoming “islands in the setting sun.”
Ensuring that aided speech outputs are placed within one’s dynamic range is a relatively easy objective to achieve without the use of any Fitting Method targets. To continue with Paul Simon’s lyric, mapping of speech is rapidly becoming the “bottom line for everyone.”
Ted Venema, PhD, teaches at Conestoga College in Kitchener, Ontario, and is the founder and director of its program for Hearing Instrument Specialists. He has a PhD in audiology from the University of Oklahoma. Ted frequently gives presentations on hearing, hearing loss and hearing aids and is author of the textbook Compression for Clinicians, published by Thompson Delmar Learning and now in its second edition.
