Longitudinal Modeling of Severe-to-profound Hearing Loss Duration and Cochlear Implant Outcomes

Introduction: Prior work has explored the impact of duration of severe-to-profound hearing loss (SPHL) on cochlear implant (CI) speech perception outcomes; however, findings have been inconsistent across studies. Duration of SPHL has largely been examined through cross-sectional correlations at discrete timepoints rather than modeled as a predictor of aural rehabilitation (AR) trajectory. This study aims to investigate the effects of duration of SPHL on rehabilitation trajectories using longitudinal modeling.

Methods: We retrospectively identified 230 CI recipients from 2015 to 2025 at a tertiary care center with available Consonant-Nucleus-Consonant (CNC) word recognition scores, known age of SPHL onset, and age of cochlear implantation. Patients with single-sided deafness were excluded, with full cohort analysis (n = 634) planned for abstract presentation. CNC scores were obtained pre-implantation and at 1-, 3-, 6-, and 12-months post-activation. Longitudinal outcomes were analyzed using a linear mixed-effects model (LMM). Duration of SPHL, time post- activation, and their interaction were modeled as fixed effects, with subject as a random effect.

Results: Duration of SPHL was not significantly associated with baseline pre-implantation CNC score. A non-significant duration of SPHL and time interaction was found at 1-month post-activation, with an estimated 0.29-point reduction in CNC score improvement per additional year of SPHL (p = 0.07). Interactions between duration of SPHL and time at later post-activation intervals similarly did not reach statistical significance.

Conclusions: Following implantation, longer duration of SPHL was not significantly associated with CNC performance at baseline or throughout post-activation follow-up. These findings highlight an important gap in the existing literature, emphasizing the need to evaluate CI outcomes through a longitudinal framework in addition to cross-sectional outcomes. They also underscore the need to further examine differences in predictive value between duration of deafness (>=90 dB) and duration of SPHL (>=70 dB) on AR trajectories.