In a column last year, Robert L. Martin described open hearing aid fittings as “the new hot topic.”1 Interest in open-fitting hearing aids has highlighted the negative effects caused by occlusion in people with hearing levels better than 40 dB HL for low frequencies.2 In the clinic, fitting these people successfully has always included striking the right balance between the amount of feedback and the amount of occlusion.3
Whenever possible, clinicians will choose to vent and then, if necessary, enlarge the vent. Often they will be satisfied with achieving a 2-mm vent as a good compromise between occlusion and feedback. But how can clinicians be confident that the vent will have the intended impact and why do occlusion problems frequently persist even after a vent is drilled?
Techniques for measuring the occlusion effect in the clinic have generally involved using the patient's own voice as the stimulus.4 While the sound pressure level in the ear canal is being measured, patients are asked to produce a sustained vowel, usually /i/ or /u/. These vowels are chosen because they are intense and have low first formants, so some of their low-frequency energy is transmitted through the solid structures of the head to the ear canal.5
By using probe-microphone measurement, the clinician can calculate the increase in the intensity of the low-frequency sound pressure in the ear canal when it is occluded. A comparison of the SPL in the canal with reference to a microphone at the entrance of the external ear canal provides a rapid quantification of the amount of occlusion produced by a particular hearing aid or earmold for a single stimulus.
There also are commercial devices available that allow this measurement to be performed quickly in the clinic.6,7 They may be more difficult to use, however, when a comparison among several different earmolds is required. To achieve a comparative measure the patient must be able to repeat accurately the sustained vowel stimulus for each measurement. Even when they can watch a sound level meter, some people find it very difficult to reproduce the stimulus vowel accurately.8,9
Revit has suggested an alternative technique.10 He proposed the use of an audiometric bone vibrator as an alternative stimulus in instances where the client is “unable or unwilling to sustain the required sound.” Because bone conduction is the dominant auditory pathway contributing to the occlusion effect,12 Revit thought that a suitable alternative stimulus might be a low-frequency pure tone via the bone conductor. Using a 250-Hz pure tone at 45 dB HL, he found a similar magnitude of occlusion effect “as when using the subject's own voice.”
In the study reported here, we found that the benefit of using the audiometric bone conductor was the accurate repetition of the stimulus, which reduces the variability of the measurement and increases the likelihood of achieving a useful comparison of different earmold conditions.
The technique described in this article may be useful in the clinic. It has been found to be a practical way to demonstrate the effectiveness of different earmold options for resolving or reducing occlusion. The findings emphasize the benefit of open fitting.
We tested 29 ears from 18 participants, 9 male and 9 female. All participants had hearing thresholds better than 40 dB HL at 250 and 500 Hz. Six participants had normal hearing and the remaining twelve had sensorineural hearing loss. The three-frequency average hearing loss for all participants was 33.5 dB HL (range 25 to 50 dB HL). The average low-frequency hearing loss at 250 and 500 Hz was 20.5 dB HL (range 2 to 32 dB). One participant was known to have been dissatisfied with his hearing aid fitting because of unresolved occlusion.
We placed a probe microphone in the ear canal of each participant. The tube curser was set to ensure that the tip of the probe tube would sit approximately 4 mm beyond the earmold canal stalk of the carved shell mold. The probe tube was connected to an Audioscan Verifit hearing aid analyzer, which was set to Speechmap for real-ear measure (REM). The Verifit stimulus was selected as speech-live and the real-ear-to-coupler difference (RECD) entered as 0 dB.
Participants wore a bone conductor on the mastoid ipsilateral to the measurement ear. We placed the bone conductor carefully so it was not in contact with the pinna, spectacles, or hearing aid case. Using a Madsen Midimate 622 audiometer, we presented a 500-Hz pure tone continuously at 60 dB HL.
We selected the 500-Hz tone because it was possible to achieve an intensity similar to a subject's own-voice stimulus. The alternative stimulus, 250 Hz, available from the audiometric bone vibrator, had a less intense maximum of 45 dB HL. Evidence from research by Stenfelt et al. suggests either frequency would produce a useful result.12 Using a similar, bone-conduction technique on a normal ear canal, they found the difference in ear canal sound pressure level for occluded and open-ear canals to be similar at 250 Hz and 500 Hz and to diminish at 1000 Hz. Maximum occlusion is likely to occur in the range of 200 to 1000 Hz,7 so whether the 250-Hz or 500-Hz frequency is selected as the stimulus, it will not coincide with the point of maximum occlusion for everyone.6
We used the Verifit's freeze curve button to capture each measurement. We measured the SPL in the ear canal for three earmolds:
- an open earmold, with the subject wearing an Oticon comfort tip;
- a vented earmold, with the subject wearing a hard acrylic carved shell earmold with a 2-mm vent; and
- an occluded earmold, with the 2-mm vent of the carved shell blocked with a putty plug.
For all measurements, the bone conductor remained in the same position. For the vented and occluded condition, only the putty was added, so the position of the probe tube was identical for the two measurements and care was taken to ensure that its position did not change for the open earmold condition. We blocked the earmold tubing by attaching a BTE hearing aid that was switched off.
The 12 participants with hearing impairment were asked to rate the perceived hollowness of their own voice for each earmold condition.
Figure 1 is an example of the resulting pattern of measurement. The change in SPL for the various earmold conditions can be easily seen. Using the graph/table button on the Verifit permits an accurate SPL to be read off the table. Using the measurements produced by this technique, we can describe the occlusion effect as the difference between the SPL in the canal for the open earmold condition and the SPL for the occluded condition at 500 Hz.
The results are summarized in Table 1. For 29 ears, the average occlusion effect (increase in the SPL in the canal at 500 Hz) was 19.5 dB (SD=4.9). The individual differences in the magnitude of the occlusion effect were large and demonstrate that some individuals have increased likelihood of finding occlusion problematic. The smallest occlusion effect measured among participants was 9 dB and the largest was 32 dB. The average reduction in SPL at 500 Hz when a 2-mm vent was added to an occluding earmold was 4.8 dB (SD=3.9; range=0 to 14 dB).
The differences among results for the vented earmold, open earmold, and occluded earmold conditions were statistically significant. A paired t test comparing the open and the occluded earmold condition gave t(28) = 20.712, p<0.001. Comparing the open earmold and vented earmold conditions gave t(28) = 14.41, p<0.001. The results of a paired t test comparing the vented earmold and occluded earmold condition gave t(28) = 6.5, p<0.001.
When a 2-mm vent was added, only 7 of the 29 ears tested had a reduction in SPL of more than 5 dB and of these 4 had a reduction of more than 10 dB. For the other 22 ears, the average impact of the 2-mm vent was 2.85 dB, and in this group of participants the average occlusion effect was 18.6 dB (Table 1).
For the 22 ears, 15.75 dB of occlusion remained at 500 Hz after venting. For 10 of the 22 ears the reduction in sound pressure with the 2-mm vent was 2 dB or less. For the 12 participants' rating of the hollowness of their own voice, a trend was found such that hollowness increased with occlusion (Figure 2).
May and Dillon reported occlusion measurement results for various earmold conditions for 10 subjects using their own voice as the stimulus.13 They found that the mean decrease in ear canal sound pressure level for 2-mm vented earmolds compared with occluding molds to be between 4 and 5 dB. This result is consistent with the finding of a 4.8-dB reduction in occlusion for a 2-mm vent for all 29 ears in this study. However, for 22 ears, the majority, the reduction was only 2.8 dB and for 10 of these ears, the reduction was less than half of the average at <2 dB.
One participant, who was always dissatisfied with the sound of his voice when aided, was of particular interest. He had a moderate, high-frequency sensorineural hearing loss and usually wore a 2-mm vent. His occlusion measurement indicated a 25-dB increase in SPL at 500 Hz when occluded. The reduction in occlusion when vented was measured at 3 dB.
The measurement technique showed that in his case the vent reduced occlusion to 22 dB, which was still above the average of 19.5 dB for the 27 occluded ears. It is no wonder that his occlusion problem had persisted. He was not a candidate for an open earmold because of feedback and neither counseling nor acclimatization time had reduced his dissatisfaction with the sound of his voice.
The simple technique described here for measuring the occlusion effect for different earmold configurations overcomes the problem of obtaining a repeatable stimulus. Although other techniques can identify people who have an increased potential for occlusion problems, tools that rely on the person's own voice require training time and have poorer stimulus repeatability, which both decrease ease of use in the clinic. Quantifying the comparative impact of earmold modifications is more difficult using these tools.5,6
The results reported here demonstrate the impact of an open earmold and an earmold with a 2-mm vent for resolving occlusion. The results highlight the dis- appointingly small reduction in occlusion for a 2-mm vent as compared with an open-earmold fitting. At 500 Hz, the impact of venting for some ears was minimal.
A huge increase in the intensity of the bone-conduction path was recorded when the open earmold and occluded test conditions were compared. However, the measurement for the vented condition was much more similar to the occluded measurement than to the open-earmold measurement.
As Dillon states: “A 2-mm vent can be regarded as a good starting point for fixing the occlusion problem, but in many cases, the vent will have to be widened to 3 mm before the patient is satisfied with the sound of his or her own voice.”14
These results suggest the importance of open fittings to reduce the occlusion effect. For the subject with persistent occlusion problems described above, open fitting is the optimal solution. However, in his case, open fitting cannot be successful without excellent feedback management. The combination of feedback management with open fitting has potential to be a much more powerful tool for overcoming occlusion than drilling out a 2-mm vent in the earmold.
We would like to thank Prof. Peter Blamey for his encouragement and assistance in writing this paper, Justin Zakis for his helpful comments and proofreading, and the 18 participants who generously gave of their time and patience while the data were collected. The research reported here was undertaken at Dynamic Hearing with the support of the Biotechnology Innovation Fund from AusIndustry (Commonwealth Government of Australia). The research was approved by the Human Research and Ethics Committee of the Royal Victorian Eye and Ear Hospital (03/529H).
1. Martin RL: Why open fittings are the next big thing. Hear J
2. Dillon H: Hearing Aids
: Sydney: Boomerang Press, 2001:130.
3. Otto WC: Evaluation of an open-canal hearing aid by experienced users. Hear J
4. Sweetow RW, Pirzanski CZ: The occlusion effect and ampclusion effect. Sem Hear
5. Killion MC, Wilber LA, Gudmundsen G: Zwislocki was right…a potential solution to the “hollow voice” problem. Hear Instr
6. Mackenzie DJ, Mueller HG, Ricketts TA, Konkle DF: The hearing aid occlusion effect: Measurement devices compared. Hear J
7. Mueller HG: There's less talking in barrels, but the occlusion effect is still with us. Hear J
8. Kampe SD, Wynne MK: The Influence of venting on the occlusion effect. Hear J
9. Shepherd D: An investigation measuring the occlusion effect with ADRO hearing aids. Unpublished masters thesis, University of Melbourne, 2004.
10. Revit LJ: Two techniques for dealing with the occlusion effect. Hear Instr
11. Reference not Provided.
12. Stenfelt S, Wild T, Hato N, Goode RL: Factors contributing to bone conduction: The outer ear. J Acoust Soc Am
13. May AE, Dillon H: A comparison of physical measurements of the hearing aid occlusion effect with subjective reports. Presented at Audiological Society of Australia Conference, Adelaide, 1992.
14. Dillon H: Hearing Aids
: Sydney: Boomerang Press, 2001:132.