Does pediatric cochlear implant insertion technique affect intraoperative neural response telemetry thresholds

Abstract Objective Recent reports of mostly adult patients suggest round window insertion is less traumatic than cochleostomy for cochlear implantation (CI), while other reports have indicated that curved electrode arrays lower the neural response telemetry (NRT) threshold and consume less power. We aimed to compare the intraoperative neural response telemetry (NRT) thresholds in children receiving cochlear implants through a cochleostomy (COCH) vs. a round window (RW) approach, as well as patients receiving a curved array vs. a straight one. Design A direct case-cohort comparison of NRT in pediatric CI recipients at two large tertiary pediatric hospitals from 2008 through 2014 was done. Univariate Mann–Whitney analyses and one-way ANOVA were performed to compare average NRT in RW vs. COCH insertion, and curved vs. straight electrodes. Multivariate regression was performed to control for age and pre- vs. postlingual patient status. Results Overall, 236 children were included, between January 2008 and October 2014 at two large tertiary referral medical centers. A total of 52 patients received a RW insertion and 184 received a cochleostomy. There was no statistically significant difference between RW insertion (187.9 ± 18.7) and COCH (183.4 ± 17.1) ( p  = 0.125). The patients were divided into four categories: RW insertion with curved electrode (175.0 ± 11.2), RW with straight electrode (192.1 ± 18.8), COCH with curved electrode (182.2 ± 16.7), and COCH with straight electrode (193.0 ± 20.8). The lowest NRT current thresholds were achieved with curved electrode array insertions through the RW ( p  = 0.001). Multivariate regression analysis revealed the following parameters were not independently associated with differences in NRT responses: RWI vs. COCH ( p  = 0.12) and pre- vs. postlingual ( p  = 0.18). The difference in NRT levels between curved electrode arrays and straight was shown to be statistically significant ( p  = 0.00075). Conclusions When controlling for insertion technique or pre- vs. postlingual hearing loss, the curved electrode array correlates with lower NRT thresholds. Although studies to examine functional language performance of these recipients are pending, initial results of this multi-institutional trial suggest that curved electrodes indeed result in lower NRT levels, particularly when inserted through the RW.