Various studies estimate the occlusion effect (OE) values differently. This book recommends the following OE values. (If using TDH earphones, the OE values are a higher.) Chapter 6 contains the rationale for use of these values.

The level of the crossed-over bone-conduction signal, as enhancement from the occlusion effect, needs to be considered in the masking formula. If you are presenting a 40 dB, 500 Hz bone-conduction signal to the right ear, while occluding the left ear with an insert earphone so that you can mask, then the level of the signal reaching the left cochlea is 50 dB (add 10 for the occlusion effect). Your masking formula will need to account for the “boost” from the occlusion effect.

Typically the audiologist will conduct unmasked bone-conduction testing on the better ear first. If you discover air-bone gaps, then . . . well, if the loss may be conductive bilaterally, then you may be running into a really challenging masking case, and potentially a masking dilemma, so you might want to plateau mask the better ear first.

Joke: Did you know that Spock (of Space Ship Enterprise fame) had three ears? His left ear, his right ear, and the final front ear. But I digress.

You might also decide to leave the better ear bone-conduction thresholds unmasked for the time being. You could plateau mask the better ear later. If the poorer ear’s hearing is significantly worse (e.g. sensorineural loss), you may not need to mask that better ear at all. The “bad” ear threshold is elevated beyond what the unmasked threshold showed. If you are sure of your masking results, then the unmasked scores must be for the better ear’s cochlea, (unless your patient is Spock and has a third ear.)

If you think the conductive loss is only in the better ear, you can formula mask when you test the poorer ear – add in both the occlusion effect and the air-bone gap size. So, let’s explore that case further, referencing Figure 9-3. As you start testing the right ear (poorer ear) by bone conduction, make the assumption that the left ear (better hearing ear) loss is conductive. (If you already plateau masked, you know exactly the size of the conductive loss.)

Figure 9-3. When calculating the needed masking in order to test right bone-conduction hearing, assume that the unmasked left ear bone-conduction scores truly reflect the cochlear sensitivity of the left ear, and use those air-bone gap sizes.

The most cautious approach is to add in air-bone gaps if you are not sure if they are or are not truly reflecting conductive involvement; so, if there is any doubt, include the air-bone gap.

Figure 9-4. Left side of figure. There is a pattern of air-bone gaps in the low frequencies, and they appear to indicate conductive involvement. When calculating the masking noise needed, one would add in the occlusion effect values (20 dB at 250, 10 dB at 500 Hz, 5 dB at 1000 Hz), and the air-bone gaps. At 2000 and 4000 Hz, it would be appropriate to add the air-bone gaps to the formula. Although the air-bone gaps are small, they appear to be “real” – it is likely that the masking noise will be attenuated 5 dB as it passes through the middle ear. Right side of figure. In this case, and especially so if patient symptoms and/or immittance testing are indicating that there is not a conductive component to the loss, adding the air-bone gaps is not necessary. They are not “real” air-bone gaps – they reflect test/re-test variability in thresholds. However, it does not usually hurt to add in these minor air-bone gaps. Ignore any bone-air gaps (e.g. 250 Hz).

If thresholds are sufficiently elevated, the responses measured may be responses to feeling the sound vibration rather than hearing the sound: they are potentially vibrotactile responses. Thresholds at or above those noted in Table 9-3 potentially may be vibrotactile.

If the patient has bone-conduction thresholds that may be vibrotactile, then what appears on the audiogram as a mixed loss may in reality be a sensorineural loss. If there is doubt – maybe the responses are hearing and maybe there is a conductive component to the loss – then treat the potentially vibrotactile thresholds as if they are hearing thresholds. Include any resulting air-bone gaps in the masking formulae.

If you are sure the responses are vibrotactile (e.g. the patient tells you that s/he felt the sound rather than hearing it; immittance testing is consistent with normal middle ear function), then the apparent air-bone gaps are of no concern. You do not need to add them to your masking formula. I would recommend marking the bone-conduction thresholds as “VT” for “vibrotactile”. (Pen in the VT notation next to the bone-conduction symbol on the audiogram. Make a comment if using a computer-based audiometer.) Vibrotactile responses will not shift with contralateral masking, so if you do mask, and the thresholds do shift, then they were not vibrotactile. (I can make an argument for a small central masking effect; the patient may be distracted from the threshold level vibration sensation by contralateral noise, but in general, thresholds of feeling aren’t going to be altered by contralateral masking.)

The concerns about overmasking are the same as for air-conduction testing. The air-conducted NOISE presented to the NON-test ear can become BONE-conducted sound that crosses BACK to the TEST ear COCHLEA. To estimate the amount of noise that could cross back, subtract 50 dB (the estimated interaural air-conduction attenuation value for insert earphones) from the masking noise level being used.

If the test ear bone-conduction threshold you are measuring is at or below (smaller number than) the crossback level, then the bone-conduction threshold is potentially elevated by the overmasking. Said differently, the reason you are measuring as high a bone-conduction threshold as you are could be because the noise crossed back to the test ear cochlea and that masking level prevented hearing of threshold level sounds. The test ear threshold is artificially made poorer because of overmasking.

It is common to need to mask bone conduction for a unilateral sensorineural loss case, so let’s begin with that case. Read the figure legends as you go along.

Figure 9-5. 500 Hz testing (10 dB occlusion effect). A. Assume that the right ear hearing is normal. The threshold was 5 dB HL both by air- and bone-conduction. If you do not think the left ear loss is conductive, then your guess is that the bone-conduction threshold will be 60 dB HL. Of course, it may be better than that, as suggested by the arrow.

B. Since there is no interaural attenuation for bone conduction, the crossover is 60 dB HL. The test frequency is 500 Hz; there is an expectation of a 10 dB occlusion effect. The level at the right ear cochlea is 70 dB HL

C. Putting in 80 dB EM into the right ear should mask the crossover, you should give a little cushion in case the patient’s occlusion effect is higher than 10 dB, or in case there is a small masking noise calibration error.

D. Assuming a 50 dB IA, the level of noise that could cross back is 30 dB HL. If the test ear bone-conduction threshold is 30 dB or better (e.g. 25 dB HL), then overmasking may be shifting the left ear bone-conduction threshold. You may be overmasking. If the loss is, as expected, sensorineural, then the presence of the 30 dB EM crossback won’t affect hearing of the 60 dB HL bone-conduction pure tone in the left ear.

Remember that overmasking happens when the crossback level is at or higher than the test ear bone-conduction threshold.

While formula masking, you need to make an estimate of the expected test-ear bone-conduction threshold. If you expected that the loss will be sensorineural, then use the air-conduction threshold in the estimating of the expected bone-conduction threshold. If, when establishing the threshold, you find the bone-conduction threshold comes in better than expected (the test ear loss is mixed or conductive), then you may need to re-adjust the masking level to a lower intensity to avoid overmasking. Figure 9-6 illustrates.

Figure 9-6. Example of 500 Hz testing (10 dB occlusion effect).

A. SNHL was expected. Masking level used = signal level (estimated threshold of 60 dB) + 10 dB OE + 10 dB pad = 80 dB EM. If you find that the bone-conduction threshold is better than expected, and may be elevated by the crossback, then you have based your formula on too high an estimated bone-conduction threshold.

B. Recalculate the minimum masking level based on the lowered estimate of the bone-conduction threshold. (30 dB expected BC threshold + 10 dB OE + 10 dB pad = 50 dB EM.) Now the cross-back (0 dB EM) is not going to affect your test results unless the bone-conduction threshold comes in at 0 dB HL or better, in which case you can reduce the masking noise further. (That’s unlikely since the unmasked bone-conduction threshold is 5 dB HL.)

As already discussed, if conductive loss is present in the non-test ear, then the NTE masking intensity needs to be increased to compensate, so that the desired level reaches the NTE cochlea. Read Figure 9-7 legends.

Figure 9-7. 500 Hz testing. The existing conductive loss is assumed to create a 10 dB occlusion effect.

A. The goal is to have 70 dB more effective masking than the NTE bone-conduction threshold (50 dB crossover + 10 dB OE + 10 dB pad).

B. Because there is conductive loss in the non-test ear, increase the masking noise by the amount of that NTE air-bone gap. This keeps the noise at the desired level when it reaches the NTE cochlea.

C. Overmasking could be a problem if the test ear bone-conduction threshold is 40 dB HL or if testing reveals even less impairment.

Conductive loss in the non-test ear increases the masking level needed. If the test ear bone-conduction threshold is relatively normal, this creates a situation where overmasking is likely, in which case one would plateau mask, if that is possible. In Figure 9-8 A, the noise level was too high, but because the bone-conduction value in the test ear was better than anticipated, the masking noise can be reduced (part B of the figure). Formula masking is still possible. Part C of the figure shows the case where the NTE ABG is large, and equal to the expected test ear ABG. In this case, formula masking is impossible. You know that you may be overmasking as soon as you begin testing.

Figure 9-8. 500 Hz testing with bilateral mixed/conductive loss. The non-test ear is assumed to have normal cochlear sensitivity.

A. The masking level was calculated assuming the loss would be sensorineural in the left ear. (50 dB expected bone-conduction threshold + 10 dB + 20 dB NTE ABG + 10 dB OE = 90 dB EM). In reality, a 30 dB HL threshold was measured. Overmasking is problematic: the crossback (90-50) may be 40 dB EM. Because the measured threshold is lower than 40dB crossback, the masking noise intensity must be decreased.

B. Recalculating the masking noise needed may allow formula masking. (30 dB or better bone-conduction threshold +10 dB + 20 dB NTE ABG + 10 dB OE = 70 dB EM). Here, crossback could be 20 dB EM. If you find the threshold is 20 dB HL or lower, you need to reduce the noise level yet again. If you are continually having to adjust your noise levels, it may be safer and easier to plateau mask.

C. When the air-bone gaps in each ear are large, then we cannot safely formula mask. When we assume that the interaural attenuation value for the air-conducted masking noise is 50 dB, we can recognize that we have the potential for overmasking. Plateau masking is needed. (Expected TE BC threshold of 0 dB crosses over + 10 dB + 40 dB NTE air-bone gap + 10 dB OE = 60. Crossback (60-50= 10 dB EM) is 10 dB EM, which is above the expected test ear bone-conduction threshold.

To be efficient, to minimize the number of times that you have to adjust the masking level, it helps to have a good guess of the test ear bone-conduction threshold, which is the first step in the Minimum Masking Level Formula. If you have prior test results, those can guide you; otherwise, here are two ways to estimate the test ear bone-conduction threshold.

Option A. If predicting normal TE BC hearing, estimate a 20 to 30 dB HL BC threshold

If you believe that the test ear loss is conductive, you are assuming that the bone-conduction threshold will be 0 dB HL. But to be efficient you don’t want to set the masking based only on that guess, because if the threshold comes in a bit higher than 0 dB HL, you will have to increase the noise level. If you think that the bone conduction scores will be normal, base your minimum level calculations on a threshold such as 20 or 30 dB HL.

Option B. Use the unmasked BC threshold as your estimate, adding 20 to 30 dB to reduce risk of undermasking

If you have already tested unmasked bone-conduction on the better ear, the test ear masked threshold won’t be better than that. Again, to reduce the number of times you need to adjust, make the calculation of the minimum masking level needed with the assumption that the test ear will have a 20 to 30 dB higher bone-conduction threshold than the unmasked threshold.

Option C. If predicting SNHL, use TE AC threshold or maximum output level

If you believe the test ear loss is sensorineural, then use the air-conduction threshold as your estimate of the bone-conduction threshold. However, if bone-conduction testing cannot be completed at that high an intensity (e.g. a 90 dB HL air-conduction threshold and a 70 dB HL bone-conduction maximum output limit on the audiometer), then you don’t need to estimate a threshold higher than what the audiometer would produce. Use the maximum output limit for bone conduction, at that frequency, when estimating the threshold – which in this case is the level at which you expect to mark “no response".

In the high frequencies, there is no occlusion effect, so just add in the conductive loss size. Remember: the safest action is to include even small air-bone gaps unless you have good reason to believe the air-bone gap is just the result of test/retest variability.

This step is critical – we must be sure we have enough masking in case the occlusion effect is larger than expected.

If you recognized the need to mask, and haven’t bothered to get an unmasked threshold for this ear, good for you! You have saved time, and your patient will appreciate your efficiency. Just obtain the masked threshold, adjusting the test ear level using the Hughson-Westlake approach.

If you have already established the unmasked bone-conduction threshold, present again at this level. If the tone is still heard, then mark the masked threshold. If the threshold is higher than the unmasked bone-conduction threshold, then use the standard Hughson-Westlake method to find the masked bone-conduction threshold.

After finding the masked bone-conduction threshold, check -- did you estimate the bone conduction threshold correctly?

If the threshold you found on your patient is at or lower than what you guessed, that is, you over-estimated the loss severity or estimated correctly, then your masking is sufficient: You have not undermasked. Mark the threshold, turn off your masking noise. Go on to the next step.

If you guessed wrong, and the threshold is higher than anticipated (e.g. you guessed a 50 dB HL threshold and the measured threshold is 60 dB HL), then you will need to make another estimate of the final bone-conduction threshold, and recalculate the masking noise that is needed based on this new guess. (Turn up the noise level.) In making the next estimate of threshold, go higher than what you have just measured – usually by 20 dB or more. The reason for increasing your guess of the test ear bone-conduction threshold by at least 20 dB is that if your estimated threshold (upon which you recalculate the masking noise) is only marginally higher than your original estimate, you are more likely to be wrong again, and have to yet again turn up the noise. If you are calculating and re-upping the masking noise continually, that is no more efficient than plateau masking.

To check for overmasking, note the noise level that you used. Subtract 50 dB – the minimum interaural attenuation value for insert earphones. If the bone-conduction threshold you just measured is at that level or lower (less impairment), then your results may have been influenced by overmasking. The noise may have crossed back to the test cochlea and masked the test ear signal. You will need to lower the noise level.

If you are potentially overmasking, and if the masking noise level was based on an estimate of the bone-conduction threshold that turned out to be too high (e.g. you guessed a 50 dB HL threshold, and you measured a 30 dB HL threshold), then recalculate the masking noise needed based on this lower threshold: try again.

If instead, the guess was pretty accurate, but you still are calculating that overmasking may be problematic, then it is likely you can’t formula mask. You will need to use a plateau masking approach in order to be confident of your test results.

If you suspect conductive or mixed loss in the TE:

• Base your MMax value on an estimate of a -10 to 0 dB HL threshold or
• If you have a prior audiogram on the patient, and have no reason to think hearing has improved, then estimate that you will find the BC threshold as 10 dB better than the last test. By selecting a value a little lower, you won’t have to adjust the masking level if, due to test/retest variability, the BC threshold is slightly better.
• If you believe the loss is sensorineural, it’s still best to guess that the threshold will come in 10 to 20 dB lower than the AC threshold. If there are minor, insignificant air-bone gaps, having based your masking formula on the lower threshold keeps you from having to lower your masking level.

With severe and profound sensorineural loss, the AC thresholds may be above the maximum bone-conduction output limits. MMax is still based on cochlear sensitivity – you can base MMax on that value rather than the maximum output of the audiometer. This will give you a very high MMax – higher than you will want to use. Adding 45 dB to the air-conduction threshold or to the maximum output of the audiometer gives you a very high MMax! Remember, clinically, don’t use a level at MMax if that will cause loudness discomfort.

To obtain MMax, add 45 dB to the anticipated bone-conduction threshold -- don’t use a masking noise level higher than this. Why 45? The interaural attenuation for air-conduction masking is 50 dB. We can safely use up to 45 dB more noise than the test ear bone-conduction threshold without concern that the crossback will interfere with the patient’s hearing of the test ear tone. As long as your established (measured) bone-conduction threshold is NOT LOWER than what you estimated as the bone-conduction threshold when making the maximum level calculation, then you will not have risked overmasking.

Remember that we must take into account the added intensity that the crossover takes on due to the non-test ear occlusion effect enhancing the “bone-conduction by air-conduction” hearing mechanism, or that the existing conductive loss creates. .

You want at least the 10 dB “safety pad”, so MMax must be at or above this calculated MMin, which includes that 10 dB safety pad. The 10 dB pad is important because some people will have occlusion effects larger than the 20/10/5 dB estimates (at 250/500/1000 Hz respectively).

The astute reader notices that, when using the MMax approach, you aren’t avoiding calculating the minimum masking level, so one may ask, why not do both calculations from the start? You could, and in cases of sensorineural loss that works well – choose a masking level between MMin and MMax. However, it’s easier to do the check afterwards. When you calculated MMax, you thought about the best possible bone-conduction threshold. When we were calculating the minimum level, we were estimating a bit higher bone-conduction threshold so that we didn’t have to readjust the masking levels. Rather than guessing the final level, the check of whether you used enough masking becomes more straight forward if you use the bone-conduction threshold that you established while using the MMax level.

If you calculated Mmax, and now find that you don’t have enough masking noise, one of two things is happening. You may have bilateral conductive loss and a potential masking dilemma. In that case, plateau mask. Another reason for undermasking is that your estimate of the test ear bone-conduction threshold was lower than the actual measured threshold. (You guessed a low threshold, the real threshold was worse than that.) In that case, your new mMax is higher. You probably can increase the masking noise intensity level.

Figure 9-10 shows an example.

Figure 9-10. It was assumed that there could be conductive loss in the test ear, so the conservative estimate of a test ear threshold of 0 dB was chosen. The MMax is 0 + 45. Let’s examine the case where a 0 dB threshold is found, as was expected. The next step is to check that the masking was sufficient. 0 dB BC threshold + 10 dB + 20 dB NTE ABG + 10 = 40. The minimum needed level is 40 dB HL, MMax is higher. Masking was sufficient. This screen is of the software program mCalc.

Let’s continue with this example, this time examining what happens when the bone-conduction threshold is not 0 dB, but turns out to be 30 dB HL. (Figure 9-11.) Continuing with the case above, the audiologist had assumed a 0 dB right bone-conduction threshold, and used 45 dB EM based on the “MMax” approach to formula masking. The right ear bone-conduction threshold was measured at 30 dB HL. The minimum level = 30 + 10 dB safety pad + 20 dB NTE ABG + 10 dB OE = 70 dB EM. The 45 dB calculated MMax is causing undermasking. As shown in Figure 9-11, with the knowledge that the test ear bone-conduction threshold is actually 30, the MMax is 75 dB EM.

Figure 9-11. When the audiologist finds a higher than expected TE bone-conduction threshold (30 dB in this case), the audiologist will need to increase the noise level. As shown in the figure, the minimum needed is now 60 dB EM. More noise can be safely used. As shown, the new MMax is 75 dB EM.

Now, let’s explore a case where the NTE does not have a conductive loss, testing 500 Hz where the occlusion effect is assumed to be 10 dB. Read Figure 9-12.

Figure 9-12. The conservative estimation is that the right ear’s (TE) bone-conduction threshold is 20 dB HL – the same as the better ear’s threshold (which was measured first, without contralateral masking). The MMax approach ensures you are not overmasking, as long as threshold is not better than you estimated. After obtaining threshold (e.g. it is as predicted at 20 dB HL), then calculate the minimum masking level, using only the occlusion effect value in this example, because there is no NTE air-bone gap. MMin is < MMax, so masking is sufficient and overmasking is not a concern.

Use the most conservative (lowest) bone-conduction threshold estimate as your basis for finding the maximum safe masking level. If you suspect conductive loss, and if you have an unmasked bone-conduction threshold, use that level. If you have not obtained the unmasked threshold but believe the loss is conductive, then use 0 dB HL (or even -10 dB HL).

If you are fairly sure that there is some cochlear loss, then think about the LEAST amount of loss that may be present, and use that as your estimate. For example, if I am retesting a patient and have the prior audiogram, with no reason to think that hearing has improved (though it may have worsened), I would choose a level a bit lower than what was found at the last evaluation. I would not use exactly the level found last time, as thresholds vary +/- 5 dB from test to retest. A level 10 dB below that found last time would be a good starting point. If I believe the loss is sensorineural, I could use the air-conduction threshold or preferably 10-20 dB below that level. (Because the MMax process uses a 50 dB interaural attenuation value, and your patient’s interaural attenuation value is likely more, it’s not imperative that you use the 10 dB below value but it’s likely to save you time. You won’t have to consider overmasking if the BC threshold is slightly lower than the AC threshold.)

If air-conduction testing has indicated a severe loss, and you anticipate that you will be recording a “no response” at the audiometer output limits, technically MMax is still based upon the cochlear sensitivity, and will yield a very high masking intensity. Subsituting the output limits of the bone-conduction transducer yields a more reasonable MMax. (Regardless of what the calculated value is, do not use a masking level that would create loudness discomfort.)

If you are wrong, and the threshold is better than that, you’ll need to recalculate. (It’s not the end of the world!)

Using your usual Hughson-Westlake threshold seeking approach, find the masked bone-conduction threshold. If you previously established the unmasked threshold, you can check for hearing at that level with masking, and if the tone is still heard, then you don’t need to re-establish threshold. However, the threshold will probably be at least 5 dB higher, due to the central masking effect.

When using the MMax approach, if the threshold is lower than expected, you may have overmasking.

Overmasking may be occurring

If BC threshold + 45 < MMax

For example, if you guessed a 20 dB HL bone-conduction threshold, which gives you the Mmax calculation of 65 dB EM, and the real threshold is 10 dB HL, you may have overmasked. 10 dB BC threshold + 45 = 55 which is lower than your MMax level of 65 dB EM. (65 dB EM can cross back to the test ear and can potentially prevent hearing of tones at or lower than 15 dB HL.) If this occurs, then use the new threshold in calculating the new, lower MMax level to use. Estimate that threshold is at least 10 dB better than what you are measuring so that, in case it was overmasking.

Another way to think about potential overmasking:

If MMax - 50 ≥ TE BC

Overmasking may be occurring

Why 45 above and 50 here? It's about less/greater than and equal to! You are OK if the noise is 45 dB above the NTE threshold; you are not OK if it is 50 dB above, or even higher intensity. Adding and subtracting even numbers is easier, so this is my preferred way of checking for overmasking.

You should always do this check, but it’s particularly important if the BC threshold is coming in higher than you guessed. After measuring the bone-conduction threshold using the MMax approach, calculate the minimum masking level with the usual formula, using the threshold you just established.

Minimum Masking Level = BC threshold + (OE + NTE ABG) + 10 dB.

Ensure that MMax was greater than or equal to the minimum level. If this is not the case, then you are undermasking. If you find that you don’t have enough masking noise, you need to use more noise to mask the crossover, or you may need to plateau mask. Sometimes you can combine a bit of plateau masking into your formula masking. If the noise level at the cochlea is only slightly higher than the TE BC threshold, then you should increase the noise slightly and check that the tone is still audible. (If not, then raising the noise either overmasked or it was previously undermasked. Remember, “If in doubt, plateau it out.”)

If the reason that you need more masking is that the bone-conduction threshold turned out to be higher than you assumed, then recalculating the MMax and finding threshold again should let you measure threshold correctly. However, if the bone-conduction threshold is similar to what you guessed, and you are observing air-bone gaps, then you can’t formula mask – you may (or may not) have a masking dilemma. Abandon the formula masking approach: plateau mask this frequency.

Which approach is best to use, Min or Max? It depends on your patient and your suspicion of site of lesion. If the patient complains of otalgia, hearing loss co-occurring with a cold, and/or has abnormal immittance findings, then you suspect conductive loss (or conductive overlay) and using the “Max” approach is more likely to work best for you.

If in the history taking your patient tells you “I can hear fine in quiet, but I’m having difficulty in noisy environments, and increasingly I have to use the phone on this ear” then you are suspecting asymmetrical, probably sensorineural, loss. The “Min” approach will likely be efficient and it will use lower levels of masking noise, which would be more comfortable for your patient. You can use MMax, but be sure not to cause loudness discomfort.

With the minimum masking approach, best for sensorineural losses, you make a guess of threshold, and you should guess that the bone-conduction threshold is as bad as it can get: i.e., equal to the air-conduction level or the audiometer maximum output level, whichever is lower. If the test ear threshold turns out to be worse than you expected, then the minimum isn’t enough. That’s why you use the “worst case scenario” (guess a high bone-conduction threshold) with the Min approach. If your guess turns out to be wrong, and the loss has a conductive element, and thus the bone-conduction threshold is significantly better than your estimate, then it is imperative that you check to see if you have overmasked. That gets us back to why you may favor MMin for the sensorineural losses, but MMax when the test ear has conductive loss.

With the MMax (maximum safe level to use) approach, best for conductive and mixed loss, you limit the noise intensity to prevent overmasking. Guess the lowest reasonably possible bone-conduction threshold. If your guess is wrong (the bone-conduction is better than even what you guessed), you’ll have to reduce the noise level. Guessing low in the first place reduces that chance. But, guessing too low increases the likelihood that your check that you have enough masking noise shows that you don’t, and you have to increase the noise level. If “max” was calculated based on too low a threshold, the true bone-conduction threshold is higher, then you may have crossover that isn’t completely masked.

You can switch back and forth between the approaches if you need to recalculate.

If in doing your MMin / MMax double checks you find that the minimum either is the same as the maximum, or worse, is higher than MMax, then it’s time to plateau mask if that is possible. For this rule to work, both MMin and MMax need to be based on the same threshold. (Normally you would guess low for MMin and high for MMax, this rule doesn’t work if you do that.)

For example, the NTE has a 25 dB air-bone gap and the TE has a 20 dB air-bone gap and a 15 dB HL bone-conduction threshold in preliminary testing. The test frequency is 500 Hz.

  MMin = 15 + 10 dB OE + 25 NTE ABG + 10 pad = 60

  MMax = 15 + 45 = 60

It would be safest to plateau, but technically you can formula mask. You are using the maximum before risking overmasking, if the interaural attenuation is 50 dB.

What if you can’t plateau? (This will come up in the next chapter – with word recognition testing, plateau is not an option.) I would advocate using the minimum level. Since the interaural attenuation is likely higher than 50, you probably won’t overmask. Annotating the results (making a note in the report) would be appropriate.

Hmmm, what would I write? “It should be noted that formula masking was used (due to… e.g. the limited attention span of the child), which is sub-optimal. Thresholds may be elevated by overmasking if the patient’s interaural attenuation is at the minimum amount clinically observed.”

Students are taught to test air-conduction at 1000 Hz first, in the better ear. That’s perfectly sensible advice. Obviously test the better ear, if there is one, first so that when you switch to the poorer hearing ear, you will know if you need to mask. Testing 1000 Hz first is good because it approximates what patients would expect to hear when you tell them to press the button when the “tone” is heard: starting at 250 Hz (which may sound like a fog horn) or 8000 Hz (a squeak, which may be inaudible with sloping sensorineural hearing loss) would potentially confuse the patient. However, once you’ve tested the patient across the range of frequencies, that confusion is removed. The patient knows what the test tones will be. So here’s a suggestion. When you test bone-conduction, START AT THE LOWEST FREQUENCY FIRST.

If the loss is sensorineural at the lowest frequency, it is likely to be sensorineural at the other frequencies as well. The “minimum masking level” approach is likely to be efficient and exposes the patient to the lowest possible noise levels. Conversely, if you find conductive or mixed loss, you may want to use the “max” approach. In this case, the low-frequency bone-conduction threshold will also help you make an estimate of the hearing at the higher frequencies. (Since hearing loss is often sloping, bone-conduction thresholds are usually the same or worse at the higher frequencies.)