Spoken word recognition in adolescent cochlear implant users during quiet and multispeaker babble conditions.

Objective: To assess overall speech intelligibility in adolescent cochlear implant speakers during quiet and multispeaker babble conditions. Study Design: A cross-sectional assessment of intelligibility incorporating group (auditory-oral versus total communication speakers), sentence context (high versus low contexts), and background conditions (quiet versus multispeaker babble). Setting: A camp designed to assess adolescents over a concentrated period. Participants: Fifty-seven adolescents who participated in an earlier study when they were 8 to 9 years old examining functional outcomes of speech perception, speech production, and language were asked to participate in follow-up study. Methods: Speech intelligibility was assessed by asking the adolescents to repeat sentences. Sentences were digitally edited and played to normal hearing listeners who either provided broad transcriptions of sound accuracy or wrote down the words they understood when the sentences were presented in quiet and in multispeaker babble. Main Outcome Variable: The dependent variables were percent correct consonants, vowels, and total words identified. Results: Very few substitutions or omissions occurred, resulting in high levels of accuracy for consonants and vowels. Speech intelligibility in quiet was significantly greater than in the multispeaker babble condition. Multispeaker babble decreased performance uniformly across sentence context for the 2 groups. Conclusion: Accurate consonant production based on measures of substitutions and omissions fails to account for distortions and allophonic variations. Reductions in speech intelligibility relative to the phoneme correct productions suggest that the allophonic variations related to distortions may influence naive listener's ability to understand the speech of profoundly deaf individuals.