The purpose of this study was to evaluate the stability of rotary chair, video head impulse test (vHIT), and vestibular evoked myogenic potential (VEMP) responses in children with normal hearing (NH) and children with cochlear implants (CIs).

Retrospective analysis of 66 children (33 males, Mage = 11.4 years, range: 3–18 years) seen in a tertiary clinic and/or research laboratory who completed rotary chair, VEMP, and vHIT across two test sessions between 2012 and 2019. The stability of these measures was compared between two groups: children with NH (n = 35) and children with CI (n = 31). For each outcome, the session difference was calculated by subtracting Session 1 from Session 2.

For rotary chair (gain and phase) and vHIT (gain), linear mixed-effects models revealed that there were no significant interactions or main effects for group (CI vs. NH), time between session, gender, or age on the session difference, suggesting that the outcomes of these measures are stable across sessions. For cervical and ocular VEMP amplitude, there was a significant interaction between group and time between sessions on the session difference. Specifically, children with NH demonstrated larger amplitudes at Session 2, whereas children with CI demonstrated smaller amplitudes at Session 2. Next, test findings were classified as normal, unilaterally abnormal, or bilaterally abnormal for Sessions 1 and 2. Misclassification was defined as a mismatch of classification between sessions. Rotary chair and vHIT had the fewest misclassifications, whereas cervical VEMPs had the most misclassifications in children with CI and ocular VEMPs had the most misclassifications in children with NH. Misclassifications in children with CI were mostly consistent with progressive vestibular loss, whereas misclassifications in children with NH were mostly consistent with improved vestibular function.

Stability and misclassification rates varied between tests and groups. Overall, rotary chair and vHIT outcomes were stable in both groups; however, VEMPs differentially changed between groups, improving in children with NH and declining in children with CI. Furthermore, despite relative stability, some children with CI evidenced progressive vestibular loss on all measures suggesting that vestibular testing should be completed serially due to the possibility of progression.