Factors Affecting Sound-Source Localization in Children With Simultaneous or Sequential Bilateral Cochlear Implants

Objectives: The study aimed to determine the effect of interimplant interval and onset of profound deafness on sound localization in children with bilateral cochlear implants, controlling for cochlear implant manufacturer, age, and time since second implant. Design: The authors conducted a retrospective, observational study using routinely collected clinical data. Participants were 127 bilaterally implanted children aged 4 years or older, tested at least 12 mo post- second implant. Children used implants made by one of three manufacturers. Sixty-five children were simultaneously implanted, of whom 43% were congenitally, bilaterally profoundly deaf at 2 and 4 kHz and 57% had acquired or progressive hearing loss. Sixty-two were implanted sequentially (median interimplant interval = 58 mo, range 3-143 mo) of whom 77% had congenital and 23% acquired or progressive bilateral profound deafness at 2 and 4 kHz. Children participated in a sound-source localization test with stimuli presented in a random order from five loudspeakers at -60, -30, 0, +30, and +60 degrees azimuth. Stimuli were prerecorded female voices at randomly roved levels from 65 to 75 dB(A). Root mean square (RMS) errors were calculated. Localization data were analyzed via multivariable linear regression models, one applied to the whole group and the other to just the simultaneously implanted children. Results: Mean RMS error was 25.4 degrees (SD = 12.5 degrees) with results ranging from perfect accuracy to chance level (0-62.7 degrees RMS error). Compared with simultaneous implantation, an interimplant interval was associated with worse localization by 1.7 degrees RMS error per year (p < 0.001). Compared with congenital deafness, each year with hearing thresholds better than 90 dB HL at 2 and 4 kHz bilaterally before implantation led to more accurate localization by 1.3 degrees RMS error (p < 0.005). Every year post-second implant led to better accuracy by 1.6 degrees RMS error (p < 0.05). Med-El was associated with more accurate localization than Cochlear by 5.8 degrees RMS error (p < 0.01) and with more accurate localization than Advanced Bionics by 9.2 degrees RMS error (p < 0.05). Conclusions: Interimplant interval and congenital profound hearing loss both led to worse accuracy in sound-source localization for children using bilateral cochlear implants. Interimplant delay should therefore be minimized for children with bilateral profound hearing loss. Children presenting with acquired or progressive hearing loss can be expected to localize better via bilateral cochlear implants than their congenitally deaf peers.