Predicting sequential bilateral cochlear implantation performance in postlingually deafened adults; A retrospective cohort study

摘要

Objective To identify which preoperative patient characteristics influence sequential bilateral cochlear implantation performance and to create a statistical model that predicts benefit. Design Multicentre retrospective cohort study. Setting All patients were operated in four academic teaching hospitals in Perth, Australia, and followed up by audiologists of the Ear Science Institute Australia. Participants A total of 92 postlingually deafened adult patients who had undergone sequential cochlear implantations between 19 June 1990 and 14 March 2016 were included. Patients were excluded if the 12‐month follow‐up consonant‐nucleus‐consonant ( CNC ) phoneme score was missing. Main outcome measure The effect of 18 preoperative factors on the CNC phoneme score in quiet (at 65 dB SPL ) with the second cochlear implant ( CI 2) one year after implantation. Results Two factors were positively correlated to speech understanding with CI 2: Wearing a hearing aid ( HA ) before receiving CI 2 ( r = 0.46, P = 0.00) and the maximum CNC phoneme score with the first CI ( CI 1) ( r = 0.21, P = 0.05). Two factors were negatively correlated: the length of hearing loss before CI 2 in the second implanted ear ( r = ?0.25, P = 0.02) and preoperative pure tone average ( PTA ) (0.5, 1, 2 kHz) before CI 2 in the second implanted ear ( r = ?0.27, P = 0.01). The following model could be created: predicted CNC phoneme score with CI 2 (%) = 16 + (44 * HA use before CI 2 (yes)) ? (0.22 * length of hearing loss before CI 2 (years)) + (0.23 * CNC phoneme score with CI 1 (%)). Because the effect of HA use before implantation played such a major role, we also created a model after exclusion of the HA factor: Predicted CNC phoneme score with CI 2 (%) = 82 ? (0.17 * length of hearing loss before CI 2 (years)) ? (0.27 * PTA in second implanted ear before CI 2 (0.5, 1, 2 kHz)) + (0.20 * CNC phoneme score with CI 1 (%)). Conclusion Advanced age or a long interval between implantations does not necessarily lead to poor CI 2 results. Patients who are successful HA users before CI 2, who have a low PTA before CI 2, a high CNC phoneme score with CI 1 and a limited length of hearing loss before CI 2, are likely to be successful CI 2 recipients.