HSC Section 8_April 2017

J. FINBOW ET AL.

MATERIALS AND METHODS

For example, a recent systematic review by Peters et al. concluded that neither the CROS nor BAHD offered much benefit for speech perception in noise (18), whereas a review by Baguley et al. suggested some benefits for both devices but more so for the BAHD (10). Improve- ments in sentence perception in noise, when noise is de- livered to the better ear, have been documented for the CROS (19) and for the BAHD (20,21). Similarly, im- provements have been reported for the BAHD when noise is delivered to the front (12,13). Both devices have been shown to impair speech perception when noise is on the side of the poorer ear (12,13,19,22); in this condition, the CROS or BAHD transmits the noise from the poorer to the better ear and thus interferes with the speech signal. Moreover, subjective benefits have been reported for the BAHD and CROS, as measured by standardized self- assessment questionnaires such as the Abbreviated Pro- file of Hearing Aid Benefit (12,16,19,21). A few researchers have examined the performance of the BAHD and CROS within the same study, thus allow- ing direct comparison between the two devices in the same sample of participants (10 Y 12,14,15,17,18,23,24). Generally, the results of these studies have favored the BAHD over the CROS based on subjective preference and outcome measures using self-assessment question- naires and on speech perception in noise. However, there are a number of methodological issues present in most of these studies (10,23). Firstly, in most studies, the CROS was always fitted before the BAHD, so there was no randomization, and in some studies only CROS fail- ures were implanted with BAHD. Only Arndt et al. (3) and Hol et al. (14) attempted randomization, but they compared the CROS to a BAHD on a headband (i.e., non- implanted), which attenuates the high frequencies (25) and can be uncomfortable. The goal of the current study is to address the need for research comparing recent models of wireless CROS hearing aids, with technologies such as digital noise reduction and directional microphones, to BAHD with similar technologies. To allow randomization while not using a BAHD headband, this study used existing im- planted BAHD users, assigned to either their BAHD or a CROS in random order for a 2-week period, and mea- sured auditory and subjective measures of perceived benefit.

Participants Of nine adult SSD BAHD users recruited, one dropped out, leaving eight participants (one male, seven female, age range 44 Y 66 yr, average 54 yr) each with BAHD experience of 0.5 to 2.5 years. The BAHD were programmed using BC direct and fine-tuned according to the patients’ comments when necessary; the devices were verified using aided soundfield thresholds. Adaptive directional microphone and noise reduction algorithm were active. None of the participants had tried CROS hearing aids before BAHD implantation. Participants’ characteristics are displayed in Table 1. Procedure Participants’ performance was compared between their own BAHD and the wireless Unitron Tandem 4 CROS hearing aid, using a within-subject repeated measures design. Data were collected over three one-and-a-half-hour visits spread 2 weeks apart. Audiometric tests were administered in a double-walled sound booth, using a Grason-Stadler GSI-61 audiometer cali- brated according to ANSI S3.6 2010 standards, with the par- ticipants sitting 1 m away from the loudspeakers. Unaided Measures Firstly, the hearing loss was measured for both air and bone conduction using insert earphones. Soundfield thresholds were then obtained with warble tones presented at 90 degrees azimuth to the better ear and at 90 degrees azimuth to the poorer ear, the difference being calculated as the head shadow. Next, unaided monosyllabic word recognition and the QuickSIN test were administered. The order of word recognition and QuickSIN testing was counterbalanced across participants. Word recognition was tested with the recorded version of the CID W-22 (Auditec of St. Louis), with a different list of 25 monosyllabic words presented at 50 dB HL in three randomized listening conditions: (a) with no noise (quiet) with words presented at 90 degrees azimuth to the poorer ear; (b) with words presented from the front (0 degree azimuth), and multitalker noise (at 45 dB HL) presented at 90 degrees azimuth to the poorer ear (S0Npe); and (c) with words presented from the front and multitalker noise (at 45 dB HL) at 90 degrees azimuth to the better ear (S0Nbe). The QuickSIN test (Etymotic Research) consists of recorded lists of six short sentences spoken by a female speaker in multitalker background noise. The multitalker noise gradually increases with each sentence presentation such that the signal- to-noise ratio decreases from 25 to 0 dB, in 5-dB steps, over the six sentences. The test measures the signal-to-noise ratio loss (SNR) with a smaller score indicating better performance. The

TABLE 1. Participants’ age, pure tone hearing threshold average, model of BAHD implanted, and length of implantation at enrollment in the study

ID Age PTA (dB HL) Better Ear

PTA (dB HL) Poorer Ear

BAHD Model

Length of Implantation Implanted Ear

1 2 3 4 5 6 7 8

49 56 44 66 46 65 57 54

6

98

Cochlear BP100 Cochlear BP100

1 yr

Right Left Left Left Left Left Left Right

10

NR NR

2.5 yr

2

Oticon Medical Ponto Pro Oticon Medical Ponto Pro Oticon Medical Ponto Pro Oticon Medical Ponto Pro

1 yr

12 11 14 15 11

93

G 1 yr

NR NR NR

1 yr

G 1 yr 2.5 yr 2.5 yr

Cochlear BP100 Cochlear BP100

56

PTA indicates pure tone average unaided thresholds for 500, 1,000, 2,000, 3,000, and 4,000 Hz; NR, no response.

Otology & Neurotology, Vol. 36, No. 5, 2015

149