Speech Intelligibility as a Predictor of Cochlear Implant Outcome in Prelingually Deafened Adults

Abstract

Among adult patients with prelingual deafness, interindividual variability in speech perception outcome after cochlear implantation is generally large. It was hypothesized that the intelligibility of the patient's own speech may be predictive of speech perception with a cochlear implant. The objectives were (1) to provide a validation of a new test battery as a measurement tool for intelligibility using a group of prelingually deafened patients; (2) to investigate the validity of the test battery as a predictor of postimplant speech perception, based on preliminary data with a cochlear implant from a subgroup of patients; and (3) to investigate the validity of the test battery as a predictor of postimplant health-related quality of life (HRQoL) for the same subgroup of patients. Twenty-five adult cochlear implant candidates with prelingual deafness participated in this study. Average age at onset of deafness was 8 mos (range 0 to 4 yrs). Speech samples from the participants were recorded and presented to two normal-hearing listeners. Results from the seven intelligibility tests in the battery were subjected to analyses of reliability and validity. Furthermore, the multiple test outcomes were submitted to a Principal Component Analysis to investigate the possibility of summarizing the data. Subsequently, from the group of 25 participants, 9 participants with above-average intelligibility were selected for implantation. Speech perception data with a cochlear implant from the nine implantees were collected at 12 mos postimplantation, as well as HRQoL data. Predictive validity of the intelligibility test battery was determined, using the postimplant data as the criterion. Results from the 25 participants averaged over listeners showed that all tests in the battery had good reliability and validity as measures of intelligibility. Principal Component Analysis showed that the multiple test outcomes could be summarized by a single underlying variable. Despite the early age at onset of deafness, the subgroup of nine participants who received a cochlear implant included several good performers in terms of speech perception with the implant. The intelligibility test battery summary score had good validity as a predictor of postimplant outcome: the more intelligible the participant's speech, the better his or her speech perception outcome with the cochlear implant. Availability of effective auditory input in early life may be the fundamental factor underlying the potential for speech perception with a cochlear implant in later life. The intelligibility test battery can be reduced to a single test to minimize testing time without negatively affecting its predictive validity. Predictive validity of the intelligibility test battery can be generalized to HRQoL outcomes of cochlear implantation, provided these outcomes are concerned with speech processing abilities. The new test battery (or its reduced version), used as a measure of intelligibility, is a promising tool for guiding cochlear implant candidacy decisions and counseling for individual patients with prelingual deafness. Because intelligibility has superior predictive power in comparison to age at onset of deafness, the latter should be discarded as an exclusion criterion for cochlear implantation.