Cochlear Implant User Affect and Reported Quality of Life

Importance: The use of patient-reported outcome measures to assess outcomes in adults who use cochlear implants has increased, as highlighted by the inclusion of the Cochlear Implant Quality of Life (CIQOL) instruments in the Minimal Speech Testing Battery, version 3. However, the self-reported nature of these instruments raises questions regarding how psychosocial characteristics impact responses. Objective: To assess whether affect and CIQOL domain scores change over time and whether affect is associated with CIQOL domain scores. Design, Setting, and Participants: Prospective longitudinal cohort study in adult cochlear implant candidates (aged 18-89 years) meeting indications for cochlear implantation based on bilateral moderate to profound hearing loss with aided sentence recognition scores 60% or less between September 19, 2019, and October 8, 2021, in a single tertiary otolaryngology referral center. Patients receiving a second cochlear implant and those without Montreal Cognitive Assessment scores were excluded. Follow-up duration was 1 year. Data analysis was performed between October 15, 2023, and August 5, 2025. Main Outcomes and Measures: Standard speech recognition tasks, the CIQOL-35 Profile, and the Positive Affect and Negative Affect Schedule (PANAS) at 4 time points were used: pre-cochlear implantation and at 3, 6, and 12 months post activation. Cohen d was used to calculate effect sizes of changes in PANAS and CIQOL domain scores. Multivariable repeated-measure mixed-effect linear models were applied to determine how positive affect and negative affect scores were associated with CIQOL domain scores. Results: Initially, 60 participants were enrolled, and 45 participants (75%) completed the study (25 female [55.6%]; median age at implantation, 67.0 [IQR, 55.0-72.0] years). From pre-cochlear implant to 12 months post activation, all CIQOL-35 domains improved. The smallest median increase was 17.7 points in listening effort domain (baseline median, 20.6 [IQR, 9.3-28.3] compared with end point median, 38.3 [IQR, 31.7-51.8]; Cohen d = 1.12; 95% CI, 0.67-1.57) and the largest was 27.8 points in the environment domain (baseline median, 31.2 [IQR, 20.6-45.1] compared with end point median, 59.0 [IQR, 48.9-65.7]; Cohen d = 1.25; 95% CI, 0.79-1.72). Positive affect increased by a median of 2 points (baseline median, 35.0 [IQR, 28.0-38.0] compared with end point median, 37.0 [IQR, 33.0-41.0]; Cohen d = 0.61; 95% CI, 0.17-1.05), and negative affect decreased by a median of 5 points (baseline median, 19.0 [IQR, 16.0-23.0] compared with end point median, 14.0 [IQR, 10.0-19.0]; Cohen d = -0.60; 95% CI, -1.04 to -0.16). In multivariable models, the regression coefficients for positive and negative affect were associated with CIQOL domain scores. For positive affect, coefficients ranged from 0.7 (95% CI, 0.4-1.0) in the communication domain to 1.4 (95% CI, 1.0-1.9) in the social domain. For negative affect, coefficients ranged from -0.8 (95% CI, -1.2 to -0.4) in the environment domain to -1.6 (95% CI, -2.1 to -1.2) in the social domain, which represented the highest regression coefficients observed across all domains and both affect measures. Conclusions and Relevance: In this study, patient affect was associated with CIQOL-35 domain scores, particularly in social and emotional areas. However, due to low regression coefficients, changes in affect were unlikely to be associated with clinically meaningful changes in CIQOL domain scores.