If a supra-aural headphone is used, e.g. a TDH series 39, 49, or 50 headphone in MX-41/AR rubber cushions, the vibration of the bony and cartilaginous areas of the external meatus, the vibration of the cartilage of the pinna, and even some vibrations of the side of the head are channeled inward towards the tympanic membrane and the occlusion effect is large. If you use an insert earphone, only the medial portion of the ear canal vibrations create sound, which is then funneled into the middle ear. Therefore, insert earphones create a lowered occlusion effect size. (Refer to Figure 2-1 for an illustration).

When you cover the non-test ear with an earphone or headphone, that causes an occlusion effect: the bone-conducted sound that has crossed over to the non-test ear has increased in intensity at the non-test ear cochlea. This means that when we start plateau masking, the initial masking level needs to be raised to compensate for the increased bone-conducted sound in the non-test ear. The question is “How much more air-conducted masking noise is needed in the non-test ear?” Obviously, that depends on the transducer used to deliver the masking noise.

In the early days of audiology, testing was conducted with TDH-style headphones. Early research fully explored the size of the occlusion effect. Common recommended values if using TDH headphones were:

250  Hz	- 30 dB
500  Hz – 20 dB
1000 Hz – 10 dB

These values represent the AVERAGE occlusion effect values. That’s interesting, given how cautious audiologists have been traditionally. Those who have a greater than average occlusion effect will have a louder sound at the non-test ear, and require a greater masking noise level. This is unlikely to be a concern if you are plateau masking where you will add in the average occlusion effect size and add an additional 10 dB that I’ll call the “safety pad” to your initial masking level. Unless the occlusion effect size were 25 dB more than expected (the 15 dB plateau width plus the 10 dB “safety pad”), one would still observe a shift in the masked threshold as the noise level increases, if the occlusion-effect-enhanced tone were crossing over. The size of the occlusion effect is more of a concern when testing using a formula approach, where one level is assumed to be enough to mask the occlusion-effect-enhanced crossover. (Formula masking is the topic of the remaining chapters.)

250 Hz – 10 dB
500 Hz – 0 to 10 dB
1000 Hz and above – 0 dB

250 Hz – 20 dB
500 Hz – 10 dB
1000 Hz – 5 dB
2000 Hz and above – 0 dB

250 Hz – 10 dB
500 Hz – 0 to 10 dB
1000 Hz and above – 0 dB.

Table 6-5. Various potentially supportable occlusion effect values are shown in this table. As illustrated, the other text recommendations and the mean occlusion effect values are substantially lower than what some patients may experience, even with deeply inserted earphones, which is why you may want to use a higher occlusion effect value.

	250 Hz	500 Hz	1000 Hz
Other texts recommend	10	0-10	0
Average of the four values	10	5	0
+1 SD approximate values	15-20	10-15	10
“Fairly cautious” (aka “proposed conservative” ) recommended values	20 dB	10 dB	5 dB

As shown in the table above, the mean occlusion effect values are a bit lower than what I recommend, particularly for formula masking where you have to anticipate the amount of sound at the non-test ear with fairly good accuracy.

The mCalc software and mQuest game that accompanies this book default to the “fairly cautious” recommended values, but allows you to put in other values if you prefer. (The audiometer simulator – AudSim – has the patient’s occlusion effect values in the patient profiles; you cannot alter them.)

In order for the occlusion effect’s bone-conduction by air-conduction enhancement to occur, the additional energy needs to pass through the tympanic membrane and middle ear. If there is significant conductive loss, then the occlusion effect is lessened or eliminated. Therefore, if testing a patient with conductive loss in the non-test ear, if that conductive loss is equal to or larger in magnitude than the occlusion effect size, you can omit the step of adding in the occlusion effect. Chapter 9 covers this concept in greater detail.

Let’s examine masking in a case of unilateral sensorineural hearing loss. Examine Figures 6-4 and 6-5

Figure 6-4. Bone-conduction masking example of testing the right ear at 500 Hz. The patient’s occlusion effect value is 10 dB. The right ear loss is sensorineural.

Figure 6-5. Bone-conduction interaural attenuation is 0 dB so the entire right ear bone-conducted signal crosses to the non-test (left) ear. When the non-test ear is occluded, the crossed over bone-conducted sound (Co) is increased to 10 dB HL (Co+OE).

To mask the right ear 500 Hz threshold in Figure 6-4 and 6-5, we would begin with the masking noise at 20 dB EM (0 dB left ear threshold + 10 dB safety pad + 10 dB occlusion effect).

Example:  
        Masking level dB EM    Threshold dB HL
            None                  0
            20                    15
                    
                    

Note that although the test ear tone is 15 dB HL, which crosses to the non-test ear at 15 dB HL, the occlusion effect enhances the tone by 10 dB, making it 25 dB HL at the non-test ear cochlea and therefore audible above the 20 dB EM. (Figure 6-6).

Figure 6-6. The 20 dB EM raised the test ear bone-conduction threshold to 15 dB HL.

In our example, the test ear true threshold is 40 dB HL, so the 15 dB HL bone-conduction signal is not yet heard in the test ear. As the masking noise is increased, threshold will increase proportionally.

        Masking level dB EM    Threshold dB HL
                25                  20
                30                  25
                35                  30  
                40                  35
                45                  40

The tone is now heard in the test ear, so the threshold will no longer increase with increases in contralateral masking noise levels.

                50                  40 "up once"
                55                  40 "up twice"
                60                  40 "up a third time- done"

Figure 6-7. The 40 dB HL bone-conducted signal to the right ear was stimulating both by the right and left ears when the masking level was 45 dB HL.

Figure 6-8. Increasing the noise to 60 dB HL prevents the cross-hearing. The threshold remains 40 dB HL because it is being heard in the test ear.

Figures 6-7 and 6-8 show that we are “on the plateau” as the noise increases from 45 to 60 dB EM. In this case of unilateral sensorineural hearing loss, the masking noise can be increased to a very high level before overmasking would occur. If the patient has an 80 dB interaural attenuation value (an average value for 500 Hz), then overmasking would occur when the noise was at 120 dB EM (80 dB above the 40 dB HL test ear threshold). See Figure 6-9.

Figure 6-9. The 120 dB EM air-conducted masking to the left ear loses 80 dB as it crosses back (Cb) to the test ear (this patient has 80 dB air-conduction interaural attenuation). The 120 dB EM would prevent audibility of the test ear tone – overmasking would occur.

Just as is true with air-conduction masking, having conductive loss in the test ear narrows the plateau width. By definition, bone-conduction thresholds are normal if the test ear has conductive loss. The patient with an 80 dB air-conduction interaural attenuation value will experience over masking once the non-test ear masking noise is 80 dB above the test-ear bone conduction threshold. (If the test ear bone conduction threshold is 5 dB HL, 85 dB EM will cause overmasking.) However, since the test ear bone-conduction thresholds are normal, you will find that you “step onto” the plateau at a low intensity level.

When the non-test ear has conductive loss, the initial masking level must be higher to overcome the loss. Since plateau masking starts at 10 dB above the non-test ear threshold, the start of the bone-conduction masking plateau will be raised, which narrows the plateau.

Just as was true for air-conduction masking, bilateral conductive loss further reduces the plateau width when testing bone-conduction. You have both the narrowing of the plateau width that is caused by starting at a higher masking level intensity due to the non-test ear conductive loss and you have the threat of overmasking when the sound crosses back to the normal test cochlea.

If you add in the occlusion effect (recall, it is not needed with significant conductive loss in the non-test ear), then you would start with higher masking levels, causing even further narrowing the plateau since it means starting masking at a higher intensity level.

Figure 6-10. An audiogram that illustrates a typical bilateral “maximum conductive loss” when testing with insert earphones.

While historically the maximum conductive loss was said to be “50-60 dB,” this was what was found when testing with supra-aural headphones. If testing with insert earphones, more loss may occur. A case where no sound is sent through the middle ear, a true maximum conductive loss, is shown in Figure 6-10, which reflects fairly average interaural attenuation values for insert earphones. (Remember that the AC signal, once it is loud enough to create skull vibration, stimulates both the test ear and non-test ear cochleas. This creates the “maximum” conductive loss.) If masking is attempted, as soon as the contralateral noise is loud enough to be audible and begins to mask the cross-over, the masking noise will cross back. Each 5 dB increase in the masking noise causes a 5 dB threshold elevation. Eventually, the results of bone-conduction testing will be no response at the output limits of the audiometer, as shown in Figure 6-11. If you initially have a loss in one or both ears that is moderately severe or worse, and then after masking have “two dead ears,” then you have experienced a classic masking dilemma. The reason the test results show no hearing when masking is that overmasking raised the thresholds.

Figure 6-11. Illustration of the results of attempting masking when there is a masking dilemma. Test results show two ears with profound sensorineural hearing loss, a result that does not make sense given that unmasked testing indicated hearing in one or both ears.

What would have to happen to have sound by-pass the middle ear entirely? A middle ear with a congenital abnormality such as being filled with bone – that would do it!

Let’s review the concept of impedance mismatch for sound transmission to the middle ear. If there were no middle ear, if the oval window functioned as the tympanic membrane, then sound would lose about 30 dB of energy because of the impedance difference of the cochlear fluid versus air. (Remember that the function of the middle ear is to increase the sound pressure through the mechanical advantage primarily coming from the collection of sound at the relatively large tympanic membrane, and the funneling of that sound into the small area of the stapes footplate.)

Sometimes conductive loss creates a situation that is worse than just the loss of the normal impedance matching transformer. Consider a break in the ossicular chain at the incudo-stapedial joint. In order to vibrate the cochlear fluids, sound has to be loud enough to vibrate the tympanic membrane which has the additional mass of the malleus and incus resting on it. The sound billowing into the middle ear is not directed just to the stapes. The sound has to be intense enough to move the stapes-laden oval window.

But even in the case of a break in the ossicular chain, the conductive loss will still allow some sound transmission inward through the middle ear. The vibrating air inside the middle ear would have to move the stapes. That’s hard to do, but it is doable.

It is typical for conductive losses not to be much worse than 50-60 dB HL, which corresponds to the old TDH earphone “maximum conductive loss.” But that’s not to say it’s a true “maximum” conductive loss.

Why this lengthy discussion? As a reminder that many times you can mask a moderate bilateral conductive loss successfully – if the loss is not truly maximum, if the air-conducted sound is not by-passing the middle ear – then you may be able to successfully plateau mask.

As mentioned previously, when bilateral conductive loss is suspected (i.e. air-conduction hearing loss bilaterally and abnormal immittance), if you are not able to plateau mask with inclusion of the occlusion effect values, try omitting them. They are not needed if the conductive loss size is equal to or greater than the occlusion effect size. This lowers the needed masking noise level, and if the loss is not quite “maximum conductive” you may be able to establish masked thresholds. The plateau width may be reduced to 10 dB, but if that plateau is repeatable, it is acceptable.

You may want to start bone-conduction testing at 4000 Hz when faced with a challenging bilateral conductive loss. Conductive losses are typically of smallest magnitude in the high frequencies, and at 4000 Hz the minimum interaural attenuation for air-conduction testing (which determines when crossback will occur) is 60 dB HL; the average is 80 dB HL, and it is possible to have a value as high as 100 dB at this frequency. Even a narrow plateau at some frequencies will help ascertain if cochlear sensitivity is asymmetrical.

Remember to document situations where your plateau width was less than 15 dB but repeatable and reliable as “reduced plateau width” test results. A reliable (repeatable) 10 dB plateau would be enough to know that you are measuring the test ear hearing; however, the additional documentation helps those reviewing the testing or trying to replicate your results.