Feasibility of a Cochlear Implant Fitting Approach Based on Phoneme Confusions: Lessons Learned from the AuDiET Study

Background: Traditional speech recognition testing in Cochlear Implant (CI) care primarily captures aggregate speech recognition performance, often overlooking detailed phoneme identification errors. This feasibility study introduces a fitting approach focusing on individual CI users' phoneme difficulties identified through self-testing paradigms.Methods: Twenty-three postlingually deaf, experienced CI users underwent fitting adjustments based on Phoneme Recognition in Quiet test outcomes. A basic fitting check was followed by advanced fitting adjustments that ranged from generic (7 out of 23) to specific adjustments targeting specific phonemes (16 out of 23).Results: The new MAP was preferred by 74% (18 out of 23) of participants, yet the aggregate phoneme identification performance showed no significant change between the pre- and post-fitting visits. However, a positive trend in targeted phoneme identification was noted (t(22) = -2.3, p = .03), approaching but not reaching conventional significance after Bonferroni-Holm correction (adjusted p = .09). A significant improvement in targeted phoneme identification was observed in the subgroup that adhered to a targeted fitting (t(11) = -3.3, P = .006, adjusted p = .03, Cohen’s d = .88).Conclusion: Using phoneme identification evaluations in the CI fitting process in experienced adult CI users is feasible.