Perimodiolar Electrode Locations Outperform Lateral Wall Arrays When Controlling for Cochlear Health and Speech...

To assess how variations in scala tympani (ST) electrode position affect speech-perception performance, controlling for cochlear health and stimulation strategy.

This retrospective cohort study included 98 postlingually deafened adult cochlear implant (CI) recipients at a tertiary referral center. Twenty-one received lateral wall electrodes (CI624; 20 mm) and 77 received perimodiolar electrodes (CI632; 18.4 mm). All implants were positioned with in the ST and programmed with the Advanced Combination Encoder (ACE) strategy. Cochlear health was quantified using pre-insertion round window electrocochleography-total response (ECochG-TR). Postoperative computed tomography with 3-dimensional reconstruction was used to determine electrode modiolar proximity (wrapping factor, WF) and angular insertion depth (AID). The primary outcome was speech perception at 6 months postactivation using CNC word scores.

ECochGTR demonstrated a strong positive correlation with CNC performance (r = 0.61; 95% CI: 0.42 to 0.84). WF correlated weakly and negatively with CNC scores (r = -0.36; 95% CI: -0.53 to -0.16), indicating better outcomes with tighter modiolar proximity. AID showed a weak positive correlation with CNC performance (r = 0.29; 95% CI: 0.08 to 0.47), with deeper insertions associated with improved scores. A regression model using ECochG-TR alone underestimated performance for tightly wrapped and deeply inserted arrays and overestimated performance for lateral wall and shallower insertions. A multivariate model incorporating ECochG-TR, WF, and AID significantly improved predictive accuracy (R² = 0.51; P = 0.001).

Electrode position within the ST varies considerably, even within the same electrode design; perimodiolar arrays may lie along the lateral wall if over inserted, whereas lateral wall arrays may approach the modiolus depending on cochlear morphology. Incorporating electrode location and cochlear health into predictive models is essential for understanding performance differences between array types. When controlling for ECochG-TR, WF, AID, and programming strategy, perimodiolar electrodes with tighter modiolar proximity outperform lateral wall electrodes at 6 months, emphasizing the importance of these factors in optimizing CI outcomes.