Bilateral Bone-Anchored Hearing Aids (BAHAs): An Audiometric

The Laryngoscope Lippincott Williams & Wilkins, Inc. © 2004 The American Laryngological, Rhinological and Otological Society, Inc.

Bilateral Bone-Anchored Hearing Aids (BAHAs): An Audiometric Evaluation Claudia Priwin, MD; Stefan Stenfelt, PhD; Go¨sta Granstro¨m, PhD; Anders Tjellstro¨m, PhD; Bo Håkansson, PhD

Objectives: Since the technique to implant bone-anchored hearing aids (BAHAs) with the use of osseointegrated implants was developed in 1977, more than 15,000 patients have been fitted with BAHAs worldwide. Although the majority have bilateral hearing loss, they are primarily fitted unilaterally. The main objective of this study was to reveal benefits and drawbacks of bilateral fitting of BAHAs in patients with symmetric or slight asymmetric bone-conduction thresholds. The possible effects were divided into three categories: hearing thresholds, directional hearing, and binaural hearing. Study Design: Prospective study of 12 patients with bilateral BAHAs. Methods: Baseline audiometry, directional hearing, speech reception thresholds in quiet and in noise, and binaural masking level difference were tested when BAHAs were fitted unilaterally and bilaterally. Results: Eleven of the 12 patients used bilateral BAHAs on a daily basis. Tests performed in the study show a significant improvement in sound localization with bilateral BAHAs; the results with unilateral fitting were close to the chance level. Furthermore, with bilateral application, the improvement of the speech reception threshold in quiet was 5.4 dB. An improvement with bilateral fitting was also found for speech reception in noise. Conclusions: Overall, the results with bilateral fitted BAHAs were better than with unilaterally fitted BAHA; the benefit is not only caused simply by bilateral stimulation but also, to some extent, by binaural hearing. Bilateral BAHAs should be considered for patients The article was presented as a poster at the American Academy of Otolaryngology–Head and neck Surgery meeting in San Diego, California, September 22–25, 2002. This work was supported in part by Swedish Research Council for Engineering Sciences (TFR 299-2000-576). From the Department of Otorhinolaryngology (C.P., G.G., A.T.), Head and Neck Surgery, Go¨teborg University, Gothenburg, Sweden; the Department of Otorhinolaryngology (C.P.), Head and Neck Surgery, Karolinska University Hospital, Stockholm, Sweden; and the Department of Signals and Systems (S.S., B.H.), Chalmers University of Technology, Gothenburg, Sweden. Editor’s Note: This Manuscript was accepted for publication July 21, 2003. Send Correspondence to Dr. Claudia Priwin, ENT-clinic, Karolinska University Hospital, 171 76 Stockholm, Sweden. E-mail: [email protected]

Laryngoscope 114: January 2004

with bilateral hearing loss otherwise suitable for BAHAs. Key Words: Bilateral BAHAs, bilateral hearing loss, hearing rehabilitation, BAHA. Laryngoscope, 114:77– 84, 2004

INTRODUCTION The hearing habilitation or rehabilitation in patients with conductive hearing loss (HL) as well as in patients with sensorineural HL combined with chronic suppurative middle ear infections or recurrent ear canal infections is difficult. The sole hearing aid solution available for these patient groups has previously been conventional boneconduction hearing aids (CBC) that work by transmitting the sound by bone conduction (BC) transcutaneously to the skull bone and to the cochlea. In Sweden, CBC has been, to a high extent, replaced with bone-anchored hearing aids (BAHAs).1 The BAHA technique by the use of osseointegrated implants was developed in Gothenburg, Sweden,2– 6 in 1977. This technique is now well established and has been in clinical use in both adults and children for 25 years. Today, there are more than 15,000 patients fitted with BAHAs worldwide.1 The vast majority of these patients are fitted with unilateral BAHAs, even though they have bilateral HL. Measurements of BC sound transmission in the skull have shown that the difference in sound transmission from one BAHA to each cochlea is normally less than 15 dB.7 Under certain conditions, especially at lower frequencies, the contralateral cochlea can receive even greater stimulation than the ipsilateral cochlea, which indicates that one BAHA should be sufficient for good hearing rehabilitation. In The Netherlands, CBCs were normally prescribed bilaterally with transducers incorporated in spectacle bows.8 The vast majority of these patients seemed to benefit from the bilateral stimulation even if it was not obvious whether this was caused by true binaural hearing or if the improved hearing was a result of greater stimulation level through the appliance of two amplifiers. Findings show that patients fitted with BAHAs bilaterally report better sound quality as well as increased ability to obtain directional hearing; they prefer bilateral BAHA fitting to unilateral.9 Priwin et al.: Bilateral BAHAs

77

The main objective of this study is to reveal benefits and drawbacks with bilateral BAHA fitting in patients with symmetric or slight asymmetric sensorineural HL in combination with a conductive impairment. Possible effects in the study have been divided in to three different categories: (1) improved hearing thresholds; (2) directional hearing; and (3) binaural hearing. In what follows, to avoid confusion, directional hearing is defined as the ability to localize the spatial direction of a sound, whereas binaural hearing is defined as the ability to use binaural cues (i.e., use the different sound information at the two cochleae to improve hearing). Moreover, bilateral stimulation is defined as the situation where sound is applied at both mastoids by BAHAs. The study includes a variety of different hearing tests such as baseline audiometry, directional hearing, speech in quiet and in noise, and binaural masking level difference (BMLD).

MATERIALS Altogether, 12 patients have been fitted with bilateral BAHAs at the Ear Nose and Throat Clinic, Sahlgrenska University Hospital, Gothenburg, Sweden; all are included in this study. Nine of the patients suffer from combined sensoneurinal and conductive HL but with no ability to use conventional airconduction (AC) hearing aids because of recurrent otorrhea or recurrent ear canal infections. The remaining three patients suffer from congenital ear canal atresia. All patients were fitted with bilateral BAHAs at least 1 year before the study, although a majority of the patients have been wearing bilateral BAHAs for several years at the time of the testing (mean time 6.8 years). Eleven of the 12 patients report daily use of bilateral BAHAs, whereas 1 patient uses bilateral BAHAs occasionally. In the following text, this report distinguishes unilateral BAHA either as unilateral BAHA on the patients’ best side (usually their first implanted side) or unilateral BAHA on the opposite side (i.e., the shadow side). Further details of the patients are shown in Table I.

METHODS The patients were equipped with two calibrated BAHAs with equal characteristics, verified by common frequency response measurement technique in practice. The BAHAs were electrically controlled by the research personnel (i.e., the BAHAs could be switched on and off by the investigator without the patient’s knowledge). Consequently, the examined patient did not know whether a BAHA was on or which BAHA was in use during the tests. Because the tests were randomized, all the patients were blinded regarding the use of unilateral or bilateral BAHAs. The volume control of the aids was preset to give maximum amplification without causing distortion of the sound levels applied in the tests. Before the tests, the type of BAHA and abutment used by the patient were registered. The patients’ best side was set and labeled as the “best” side throughout the tests. The best side was usually equal to the aid first implanted.

Tone Thresholds Pure AC and BC tone thresholds were measured to obtain an accurate hearing status for the patient by using standard audiometric procedures and equipment (Interacoustics AC-40, Assens, Denmark). For measurement of AC thresholds, TDH-39 earphones were used and for the BC thresholds, a Radioear B-71 (Telephonics Corp., Farmingdale, NY) bone transducer (Radioear Corp., New Eagle, PA). The measurements were performed in a sound insulated room with a background noise of less than or equal to 22 dBA during the tests. The earphones and bone transducer were set according to International Standards Organization (ISO) 389 and ISO 7566, whereas the sound field was set according to ISO 226 and International Electrotechnical Commission (IEC) 645–2. The examined patient was accommodated in the sound-insulated room, with an arrangement of 12 loudspeakers spaced at 30° intervals. A complete illustration of the test setup is shown in Figure 1. Speakers were placed in a circle, with a 1 meter radius from the patient and at a height equivalent to the head of a sitting patient. Current speaker setup was used for the free sound-field tests. The tone thresholds were measured in

TABLE I. Characteristics of the 12 Included Patients. BAHA Experience

Number

Sex

Age (years)

Type of BAHA Left

Type of BAHA Right

Abutment Left

Abutment Right

First Fitted Side

Subjective Best Side

Uni (years)

Bil (years)

First Side AC (dB HLac)

Second Side

BC (dB HLbc)

AC (dB HLac)

BC (dB HLbc)

Etiology

1

M

42

Comp

Comp

S

S

L

R

14.0

6.7

55

17

27

20

CO

2

F

68

C300

C300

B

B

R

R

17.5

15.1

87

52*

102

45

CO

3

F

55

C300

C300

B

B

L

L

15.4

8.6

48

18

53

27

CO

4

F

60

C300

C300

B

B

R

R

10.4

2.6

55

35

62

33

CO

5

M

58

Comp

Comp

S

S

R

R

16.6

3.0

82

53

70

50

CO

6

F

65

C300

Comp

B

S

R

R

16.9

2.9

73

43

68

43

CO

7

F

65

C300

C300

S

S

L

L

10.7

9.7

50

28*

77

47

CO

8

F

51

Comp

Comp

S

S

R

L

5.8

1.0

62

40

28

25

EO

9

F

47

C300

C300

S

S

L

L

10.9

1.4

42

27

58

27

CO

10

F

52

C300

C300

B

S

L

R

21.0

19.6

58

27*

67

32

CA

11

F

27

C300

C300

S

S

R

R

15.2

1.0

50

10

48

7

CA

12

M

30

Comp

Comp

B

S

R

R

17.0

10.1

38

8

48

15

CA

For each patient, sex, age at the time of the study, type of BAHA at each side (Comp ⫽ Compact; C300 ⫽ Classic 300), type of abutment on each side (S ⫽ snap coupling; B ⫽ bayonet coupling), first fitted side (L ⫽ left; R ⫽ right), subjective best side, experience with unilateral and bilateral BAHA fitting, pure-tone average thresholds of the frequencies 0.5, 1, and 2 kHz for air conduction (AC) and bone conduction (BC), and etiology are given. Thresholds are given for both first and second fitted side. Patients 1, 2, 3, 4, 5, 6, 7, and 9 are diagnosed with chronic otitis (CO); further underlying etiology is unknown. Patient 8 has recurrent external otitis (EO) in combination with otosclerosis. Patients 10, 11, and 12 have congenital atresia (CA). In patient 12, the atresia is a part of the Treacher Collins syndrome. *At at least one frequency with no fixed value was attained because of distortion; the value included for the mean was the highest value measurable.


78

Priwin et al.: Bilateral BAHAs

ing loudspeakers simultaneously (surrounding noise). The test was conducted with unilateral BAHA on the patient’s best side and with bilateral BAHAs by using one practice list and two measurement lists (i.e., a total of 6 practice lists and 12 measurement lists).

Binaural Masking Level Difference Test

Fig. 1. Test setup used during the study. Three different boneanchored hearing aid (BAHA) conditions were tested: unilateral BAHA on the best side, unilateral BAHA on the shadow side, and bilateral BAHAs.

the sound field using warble tones recorded and presented using the Be´ ke´ sy sweep method. The free field tone thresholds were tested at four different directions: in front (0°), at the right and left side (⫾90°), and from the behind (180°).

Directional Hearing Directional hearing was evaluated with the same speaker setup as for the tone thresholds. A narrow-band (1/3 octave) noise centered at either 0.5 or 2.0 kHz was presented at 65 dB HL for a 1 second duration. The directional hearing ability was tested for three different BAHA options: (1) unilateral BAHA on the patient’s best side; (2) unilateral BAHA on the shadow side; and (3) bilateral BAHAs. The stimuli were presented three times from each speaker and for each option, according to a randomized sequence of three presentations from each speaker: ⫻3 BAHA conditions ⫻ 12 speakers ⫻ 2 frequencies ⫽ 216 tests.

A test that has shown to be sensitive to proving the existence of binaural hearing is the BMLD test. This test is carried out with bilateral BAHAs. A signal (pure tone) is presented in noise, and the task is to detect the tone. Three different conditions are tested. For the first condition, the same signal and noise are presented equally at both ears; this condition is denoted S0N0. The following condition is when the phase of the tones presented at the two sides has an opposite phase, but the noises are in phase (the levels are equal at both sides). In other words, the tone at the left side is 180° out of phase compared with the tone at the right side; this condition is denoted S␲N0. The third condition is when the noises at both sides are 180° out of phase, but the tones are in phase, denoted S0N␲. The audio inputs of the BAHAs were used during the BMLD test, which means that the signal (both tone and noise) is supplied directly from the audiometer to the transducer of the BAHA, circumventing the microphone. The test was conducted using a signal at three different frequencies, 0.25, 0.5, and 1.0 kHz, combined with a narrow band noise centered on the corresponding signal frequency, at a level of 65 dB HL. In total, 18 estimations were performed: 3 frequencies ⫻ 3 phase conditions ⫻ 2 repeats ⫽ 18.

RESULTS Baseline Audiometry Nine of the 12 patients had mixed HL, and the remaining 3 had primarily conductive HL. Of the patients with primarily conductive HL, two patients had congenital ear canal atresia with almost normal cochleae. The third patient with primarily conductive HL had recurrent otorrhea, which at times gave a temporary conductive HL. For 10 of the 12 subjects, the BC thresholds were symmetric (i.e., the pure tone BC thresholds averaged at 0.5, 1, 2, and 4 kHz did not differ more than 10 dB between the ears), whereas the bone-conduction thresholds of the other two patients were partially asymmetric. However, the BC thresholds at the two sides of the asymmetric patients did not differ by more than 20 dB. For more detailed information of the patients see Table I.

Speech Reception Threshold Speech reception thresholds were measured in both quiet and in noise with phonetically balanced three-word sentences. These three-word sentences were extracted from the five-word sentences developed by Hagerman10 (i.e., the first 2 words of each sentence were removed). Each test list was composed of 10 threeword sentences. The sentences were presented by a female voice, with the aim of finding the noise level giving a 50% correct score. Altogether, three lists were presented for two BAHA options: unilateral (on the patient’s best side) and bilateral BAHAs. The first list was used as a practice list, whereas the test was conducted twice (i.e., a total of 2 practice lists and 4 measurement lists). The speech was presented at the loudspeaker in front of the subject (0°). The list sequence was randomized among the patients. The speech reception threshold in noise was tested with the speech level at the patients most comfortable level (between 65 and 80 dB HL), presented at 0°. The noise was speech weighted, with the presentation at either ⫹90°, ⫺90°, or from all 11 remain-


Free Sound Field Tone Thresholds Free sound field tone thresholds from the four directions 0°, ⫹90°, ⫺90°, and 180°, were tested with unilateral and bilateral BAHA fitting. Average improvements of the thresholds with bilateral fitting compared with unilateral fitting are shown in Figure 2. While the sound was presented in front, at the best side, and from behind of the patient, the average improvement with bilateral fitting was between 2 and 7 dB for the investigated frequency range 0.25 to 8 kHz. This improvement is similar to an energy doubling (corresponds to an improvement of 3 dB) and a double increase of signal amplitude in phase (corresponds to an improvement of 6 dB). When the sound was presented at the shadow side, the average improvement caused by bilateral fitting was greater: between 5 and 15 dB. In this case, the improvement caused by bilateral Priwin et al.: Bilateral BAHAs

79

microphone in the unilateral case; with bilateral fitting this attenuation is removed, and, consequently, better hearing is achieved. The results differed greatly among the examined patients. The standard deviation (SD) of the improvement with bilateral fitting in Figure 2 was almost as great as the average improvement. Consequently, the SD of the improvement when the sound was presented at the shadow side was greater than for the other three sides.

Directional Hearing

Fig. 2. The improvement of free field tone thresholds without bilateral bone-anchored hearing aid (BAHA) fitting compared with unilateral BAHA fitting. The four curves show the result when the stimulation is presented in front (0°), at the right side (90°), at the left side (⫺90°), and from behind (180°). The results are the averages values for the 12 included patients.

BAHA fitting can be explained by the head-shadow effect. Normally, the air-borne sound from the shadow side is attenuated by the head before it is picked up by the BAHA

Directional hearing was evaluated by the patient’s ability to locate a sound source. The presented sound was given from one of the 12 loudspeakers placed in a circle by 30° of interval. The stimuli were presented three times from each direction, altogether 12 directions, centered around a frequency of 0.5 or 2 kHz, and for three BAHA options: (1) BAHA on the left side; (2) BAHA on the right side; and (3) with bilateral BAHAs. The results from the directional hearing test are shown in Figure 3 where, for a given BAHA condition, the response angle is plotted in relation to the stimulation angle. The larger the circle in the plot, the more responses for that specific response angle. Figure 3a shows the result for a given stimulation angle when the sound is centered around 0.5 kHz and the BAHA is fitted on the patient’s left side. In this case, most sounds are perceived by the patient as coming from the

Fig. 3. The results from the directional hearing test of the 12 patients. The test was conducted with a single-side bone-anchored hearing aid (BAHA) and bilateral BAHAs at frequencies centered at 0.5 and 2 kHz. The horizontal axis representing the loudspeaker angle and providing the stimuli and the vertical axis is the perceiving location by the patient. The dots are separated for better visualization. The diagonal line indicates the correct responses.


80


left side and slightly from behind. A similar result is found for 0.5 kHz and with the BAHA on the right side (Fig. 3b); most sounds are perceived as coming from the right side and slightly from the back. With bilateral BAHAs, the sound localization is better (Fig. 3c); more of the results are centered on the diagonal line that, in this case, represents the correct answer area. If a patient has a perfect direction hearing ability, all responses should be on or in close conjunction with the diagonal line. Figure 3f presents the results when the stimulation is centered around 2 kHz. These results are similar to the results when the stimulation is centered around 0.5 kHz (Fig. 3, a to c); with the BAHA on the left side, the response is lateralized toward the left and back, with the BAHA on the right side, the response is lateralized toward the right and back, and with bilateral BAHAs, the result follows the diagonal line better. A different way to present the results from the directional hearing test is presented in Figure 4. The results are presented as either the correct score or whether the response is within 30° of the stimulation angle. All scores are presented for the following three BAHA conditions: (1) unilateral BAHA on the patient’s best side, (2) unilateral BAHA on the patient’s shadow side, and (3) with bilateral BAHA fitting. With unilateral BAHA, either placed on the best side or on the shadow side, the results are close to the chance level (8.3% for correct and 25% for within 30°). When bilateral BAHAs are used, there is a significant increase in the ability to localize the sound source.

Speech Reception Scores Speech recognition in quiet was measured with speech presented in front of the patient (0°). The test was conducted using unilateral BAHAs at the patient’s best side and with bilateral BAHAs. The average threshold for speech in quiet with unilateral BAHA was 38.7 dB HL, whereas with bilateral BAHAs, it was 33.3 dB HL. Consequently, the improvement with bilateral fitting is 5.4 dB. This improvement was significant (P ⫽ .001) when tested with the paired t test.

During the speech in noise tests, the speech was presented in front of the patient and the masking noise at either the left or the right speaker (⫾90°), or from all the remaining 11 speakers simultaneously (surrounding noise). When the masking noise was presented at the best side, the difference in signal-to-noise ratio (SNR) threshold between bilateral and unilateral fitting showed an improvement of 3.1 dB (i.e., the SNR threshold with bilateral fitting was 3.1 dB lower than with unilateral fitting). When the masking noise was presented from the shadow side, with bilateral BAHA fitting, the speech reception thresholds deteriorated compared with unilateral fitting with 1.0 dB. In this particular case, the SNR thresholds with bilateral fitting were 1.0 dB higher than those obtained with unilateral fitting. When the masking noise was presented as surrounding noise (i.e., from the remaining 11 speakers surrounding the patient), speech reception threshold was improved by lowering the SNR threshold by 2.8 dB when testing with bilateral BAHAs. All results were obtained with the speech signal at the patient’s most comfortable level.

BMLD Measurements The results of the BMLD measurements are shown in Figure 5. The tests were conducted at the frequencies 0.25, 0.5, and 1 kHz for three test conditions: (1) signal and noise in phase at both sides (S0N0), (2) signal 180° out of phase and noise in phase (S␲N0), and (3) signal in phase and noise 180° out of phase (S0N␲). Averages and individual results are given in Figure 5 as changes from the S0N0 condition (i.e., all results are individually normalized as 0 dB for the S0N0 condition). At 250 Hz, the changes for the different conditions were not great: for S␲N0, the changes of the thresholds were within 3 dB except for two patients, and for S0N␲, the thresholds changes were between ⫺18 and 3 dB, with an average of ⫺5 dB. However, results obtained at 0.5 and 1 kHz were similar for both S␲N0 and S0N␲: the thresholds changes were between ⫺10 and 10 dB except for a couple of the patients. At 0.5 kHz, the average threshold changes were 2 dB for S␲N0 and ⫺4 dB for S0N␲, and at 1 kHz, the average threshold changes were 3 dB for S␲N0 and ⫺3 dB for S0N␲.

DISCUSSION

Fig. 4. The average results from the 12 patients during the sound localization test. The result is presented as percentage correct answers and answers within 30° of the correct response, where the bone-anchored hearing aid (BAHA) fitting is on the best side, shadow side, and with bilateral fitting. The test is conducted with the stimuli centered on 0.5 and 2.0 kHz. The chance levels for a correct answer (8.3%) and an answer within 30° of the stimulation direction (25%) is indicated with dashed lines.


Approximately 750 adults and 40 children have been fitted with BAHAs at the Ear, Nose and Throat Clinic, Sahlgrenska University Hospital. Of these, only 12 patients have been fitted with bilateral BAHAs, all of them adults. In a study by Markides,11 it was clearly shown that many patients with symmetric HL prefer bilateral amplification when fitted with AC hearing aids. This may also be true for BAHA patients with symmetric or slight asymmetric bilateral sensorineural HLs in conjunction with a conductive loss. The patients at our clinic already fitted with bilateral BAHAs report subjectively better hearing with two BAHAs than one, and they are commonly using bilateral BAHAs daily in most hearing situations. However, in noisy surroundings with a dominant noise source, they usually turn off one of the BAHAs. The current study was conducted to evaluate whether objective hearing improvement was obtained with bilateral BAHAs compared Priwin et al.: Bilateral BAHAs

81

Fig. 5. Individual (thin curves) and average (thick curves) results of the binaural masking level difference (BMLD) test. The results are presented as the relative threshold change of S兿N0 and S0N兿 compared with S0N0 at the frequencies 0.25, 0.5, and 1.0 kHz.

with unilateral fitting; this study evaluates the aspects of directional hearing, speech reception thresholds in quiet and in noise, and BMLD. To date, only a few studies have been conducted on patients fitted with bilateral BAHAs12–16. Hamann et al.15 preformed the first study in 1991; in that study, speech reception thresholds in quiet were tested. These results showed an improvement of the speech reception thresholds in quiet, with 4.0 dB when bilateral BAHAs were used. Similar results have been reported by Bosman et al.12 The tests in this study included directional hearing, speech reception thresholds, and BMLD on 25 patients fitted with bilateral BAHAs. Altogether, their results show that patients fitted with bilateral BAHAs obtain better directional hearing, and they also gain in improved speech reception thresholds in quiet and in noise. They further found BMLD results indicating binaural hearing with bilateral BAHAs. In the current study, one of our aims was to investigate whether previous results were reproducible. Therefore, similar hearing tests were conducted with some smaller alterations to gain some extra information, especially about directional hearing.

Directional Hearing In previous studies concerning bilateral BAHAs, directional hearing has been tested with loudspeakers positioned in a semicircle at 30° intervals around the patient. In this study, the speakers were placed at 30° intervals in a full circle. This setup was used to test the patient’s ability to differentiate between left- and right-side stimulation as well as between front and back stimulation. Consequently, it is harder for the patients to give a correct answer in the tests with the full-circle strategy than with the half-circle strategy. This can explain why the directional hearing results in our study show slightly lower values than in the previously mentioned studies. However, the sound localization ability was significantly better with bilaterally fitted BAHAs than with unilateral fitting. The results with unilaterally fitted BAHA were close to Laryngoscope 114: January 2004

82

the chance level, whereas the results with bilateral fitting were significantly better than the chance level. With unilateral fitting, the difference between the patient’s two sides was not of significant importance. It is generally assumed that at low frequencies (below 1 kHz) the interaural phase difference is the major cue to determine the sound location. At high frequencies on the other hand, the major cue is the interaural level difference. The results from the directional hearing test (Figs. 3 and 4) show similar results at both 0.5 kHz and 2 kHz with bilateral fitted BAHAs. This indicates that the patients were able to use time differences as well as level differences between the two stimulation positions to estimate the sound direction.

Speech Reception The average improvement with bilateral BAHA fitting for speech reception in quiet was 5.4 dB. This gain is likely a result of two contributing processes: energy summation at the cochlear level and diotic summation at the central auditory level. With the sound presented at the front of the patient, the sound pressure at both BAHAs is approximately equal. The vibration applied at the mastoid is transmitted to both cochleae, with minor differences in amplitude and phase. Therefore, with two stimulation positions (bilateral BAHAs), a general energy increase at both cochleae is expected, and, consequently, there is some lowering in threshold levels. Furthermore, the transcranial attenuation typically increases with frequency.7 In general, this means that the high-frequency sound is greater at the ipsilateral cochlea in comparison with the contralateral cochlea. Therefore, the improved highfrequency hearing is a result of perception at the central auditory level. Bilateral BAHA fitting also improved the patients hearing in noisy environments. When the noise source was placed at the unilateral fitted side (best side), the SNR threshold with bilateral fitting improved with ⫺3.1 dB. This is understandable because both BAHAs receive equal amounts from the sound signal (source at the front), Priwin et al.: Bilateral BAHAs

but the BAHA on the shadow side has a greater SNR because the head attenuates the noise. When the noise source is on the shadow side, the SNR threshold deteriorated with 1.0 dB when bilateral BAHAs are used. This can be explained by the increase of noise transmitted to the ears with an extra BAHA on the shadow side (noise side). The third condition for measuring speech reception in noise was established by presenting the speech signal at the front of the patient and the noise from the remaining 11 speakers. In this case, the noise can be considered to be a diffuse field. The improvement of the SNR threshold with bilateral BAHAs was 2.8 dB. These results indicate that if there is a dominant noise source, the patient should only use one BAHA, provided the noise source is placed on the shadow side. If it is not possible to have the noise source on the shadow side, if the noise is caused by multiple sources, or if it is a diffuse noise field, bilateral BAHAs are beneficial for the patient.

BMLD If a signal and noise source are spatially separated, the signal can be detected with worse SNR with binaural hearing than with monaural hearing. This means that the phase relation between the signal and noise differs at the two ears. The best detection ability is when the phase relation of the signal and noise differs 180° between the ears. This can be tested with the BMLD test. When the signal and noise are presented at same phase relation to both ears, no extra information is provided. However, if the signal or noise phase is altered at one ear, the signal can be detected at a worse SNR, which is termed release from masking. In the BMLD test conducted in the study, three different phase relation conditions were tested: (1) signal and noise in phase at both sides (S0N0), (2) signal 180° out of phase and noise in phase at the two sides (S␲N0) and (3) signal in phase and noise 180° out of phase at the two sides (S0N␲). Compared with S0N0, detection threshold for the signal can be 15 dB lower for S␲N0 and S0N␲ with binaural hearing for frequencies below 1.5 kHz. It should be remembered that the BMLD is derived for AC testing. With BC testing, there is a significant amount of transcranial transmission that is not present in the AC case. This means that the signal and noise at one side is affected by the phase and amplitude of the bilateral stimulation. The amplitude is equal at both sides for all three conditions, but the phases differ. The signal is a tone that adds either constructively or destructively. The noise is a broadband signal that can be considered to add as energy (i.e., the masking level at the cochlea is approximately independent of the phases of the noises). This means that, with the addition of stimulation from the two sides at the cochlear level, S0N0 and S0N␲ should give similar signal detection thresholds. The difference between S0N0 and S␲N0 at the cochlear level depends on the phases of the transmission between the mastoid and the ipsilateral and contralateral cochlea. Figure 5 shows that there are large individual differences in the BMLD results, but the general trend is a spread of the result for S␲N0, which indicates signal interaction, both in phase and out of phase; the average is Laryngoscope 114: January 2004

⫺2 to 3 dB for the three frequencies. S0N␲ also shows large individual differences, but with a general trend toward lower thresholds: the average thresholds for the three frequencies are ⫺3 ⫺5 dB. These results, as well as the results from the hearing localization test and speech in noise test, indicate that bilateral fitting of BAHAs not only gives bilateral stimulation of the sound but also, to some extent, gives binaural hearing. However, because of the cross-over transmission between the two cochleae with BC stimulation, the binaural effects are expected to be less than with AC stimulation. For almost all the tests conducted, the congenital ear malformations subgroup (3 patients) performed better than the chronic otitis subgroup (9 patients). Although it is hazardous to draw conclusions from such small groups, the difference between the two groups originates most probably in the general better sensorineural hearing for the congenital ear malformation group. Because of the better sensorineural hearing, the patients in the congenital ear malformation group are able to better use the binaural cues that seem to exist for bilateral BC stimulation. One of the objectives of the study was to evaluate drawbacks and within-research-scope benefits with bilateral BAHAs. On the basis of the hearing tests conducted in the study, we could not show any drawbacks with bilateral BAHAs except for a deterioration in SNR in speech in noise test if noise was presented at the shadow ear. The results of the other hearing tests conducted were all in favor of bilateral BAHAs. In this study, we have compared bilateral with unilateral BAHA fitting from an audiologic point of view and have not done any close research on the cost benefit or the extra effort it takes for both doctor and patient with two implants. Finally, the study shows benefits far greater than drawbacks for bilateral BAHAs.

CONCLUSION Twelve patients with bilateral BAHAs were tested in the study. They all had symmetric or slight asymmetric sensorineural HL and an additional conductive component. Of these patients, 11 were using bilateral BAHAs on daily basis. Audiologic tests showed that the patients benefited from bilateral fitted BAHAs compared with unilateral fitted BAHAs at the areas of tone thresholds, sound localization, and speech reception thresholds in quiet as well as in noise. Moreover, the release from masking found through the BMLD test gives clear indication that bilateral BAHA fitting provides some binaural stimulation. The major benefits of bilateral BAHA fitting should therefore be offered to patients with symmetric sensorineural HLs otherwise suitable for BAHAs.

Acknowledgments The authors thank Ann Edensva¨ rd, audiologist.

BIBLIOGRAPHY 1. Tjellstro¨ m A, Håkansson B, Granstro¨ m G. Bone-anchored hearing in adults and children. Otolaryngol Clin N Am 2001;34:337–363. 2. Tjellstro¨ m A, Lindstro¨ m J, Halle´ n O, et al. Osseointegrated


83

3.

4. 5. 6.

7. 8.

9.

titanium implants in the temporal bone. A clinical study on the bone anchored hearing aids. Am J Otol 1981;2: 304 –310. Håkansson B, Tjellstro¨ m A, Rosenhall U, Carlsson P. The bone anchored hearing aid: principle of design and a psychoacoustical evaluation. Acta Otolaryngol (Stockh) 1985; 100:229 –239. Håkansson B, Carlsson P. Skull simulator for direct bone conduction hearing devices. Scand Audiol 1989;18:91–98. Håkansson B, Carlsson P, Tjellstro¨ m A. The mechanical point impedance of the human head, with and without skin penetration. J Acoust Soc Am 1986;80:1065–1075. Tjellstro¨ m A, Rosenhall U, Lindstro¨ m J, et al. Five year experience with skin penetrating bone anchored implants in the temporal bone. Acta Otolaryngol (Stockh) 1985;95: 568 –575. Stenfelt S, Håkansson B, Tjellstro¨ m A. Vibration characteristics of bone conducted sound in vitro. J Acoust Soc Am 2000;107:422– 431. Cremers CWRJ, Snik AFM, Beynon AJ. Hearing with the bone-anchored hearing aid (BAHA, HC 200) compared to a conventional bone-conduction hearing aid. Clin Otolaryngol 1992;17:275–279. Dutt SN, McDermott AL, Burrell SP, et al. Speech intelligibility with bilateral bone-anchored hearing aids: the Bir-


84

10. 11. 12. 13. 14.

15.

16.

mingham experience. J Laryngol Otol Suppl 2002;28: 47–51. Hagerman B. Sentences for testing speech intelligibility in noise. Scand Audiol 1982;11:79 – 87. Markides A. Binaural Hearing Aids. New York: Academic Press, 1977. Bosman AJ, Snik AF, van der Pouw CT, et al. Audiometric evaluation of bilaterally fitted bone-anchored hearing aids. Audiology 2001;40:158 –167. Snik AF, van der Pouw CT, Beynon AJ, et al. Binaural application of the bone-anchored hearing aid. Ann of Otol Rhinol Laryngol 1998;107:187–193. van der Pouw CT, Snik AF, Cremers CW. Patients’ opinions about bilateral BAHA application compared to unilateral application. In: Bone Anchored Hearing. Short and long term results. Dissertation by CTM van der Pouw, University of Nijmegen, The Netherlands 1998: 129 –139. Hamann Ch, Manach Y, Roulleau P. La prothe`se auditive a´ ancrage osseux BAHA (bone anchored hearing aid). Resultats applications bilaterales. Rev Otolaryngol 1991;112: 297–300. Dutt SH, McDermot AL, Burrell SP, et al. Patient satisfaction with bilateral bone-anchored hearing aids: the Birmingham experience. J Laryngol Otol Suppl 2002;28: 37– 46.
