CAN ARTICULATION INDEX BETTER ESTIMATE SPEECH INTELLIGIBILITY ON ADULT COCHLEAR IMPLANT USERS COMPARED TO AIDED PURE TONE THRESHOLDS?

Abstract

Introduction and objective: Aided pure-tone audiometry is often performed on cochlear implant (CI) users to evaluate speech sound accessibility. This study examines the relationship of speech recognition thresholds (SRT) of CI users using 1) Bisyllabic Malay Speech Audiometry (BMSA) and 2) Malay Matrix Sentence Test (MMST) with pure-tone audiometry aided thresholds (PTA_AT) as well as the articulation index (AI). Methods: In this cross-sectional study, SRT measurements for all three speech tests were collected from nineteen (average device age of 4.2 ± 3.7 years) post-lingual adult CI users. Participants had a median age of 37 years old (IQR = 17.5) and PTA_AT of 34 dB HL (IQR = 5.5). Results: Median SRT of BMSA and MMST were 45 dB SPL (IQR = 7.5 dB) and -4 dB SNR (IQR = 4.9 dB SNR), respectively. Spearmen’s rank-order correlation revealed no statistically significant correlations between average PTA_AT and the SRT of BMSA (r_s(19) = 0.396, p = 0.09) and MMST (r_s(19) = 0.135, p = 0.582). Spearmen’s rank-order correlation also revealed no statistically significant correlations between average AI and the SRT of BMSA (r_s(19) = -0.169, p = 0.489) and MMST (r_s(19) = 0.035, p = 0.887). Conclusion: Both PTA_AT and AI are poor estimators of speech perception abilities with and without competing noise. Speech tests should be routinely performed on CI users as neither aided thresholds nor AI are reliable measures of speech-sounds accessibility.