Reducing the number of impulses in video head impulse testing - It's the quality not the numbers.

ABSTRACT

OBJECTIVES: The video-Head-Impulse-Test (vHIT) is widely used to evaluate vestibular function. Nevertheless, there is no consensus on the necessary or ideal number of impulses performed for robust VOR gains. Therefore, the aim of our study is to analyze how many impulses are needed to receive reliable VOR gains in difficult testing situations like testing children younger than 48 months and in children with vestibular loss. METHODS: A retrospective data analysis was performed in which existing vHIT results of 25 healthy children aged 5-48 months were included as well as vHIT results of 25 children with vestibular loss aged 2-16 years. Descriptive data analysis was performed and further statistical analysis was conducted to determine if the number of head impulses could be reduced using internal consistency (Cronbach's alpha) and paired t-test. RESULTS: Median gain was 0.95 (±0.16) for impulses to the right and 0.97 (±0.16) for impulses to the left in healthy children and ranged from 0.01 (±0.11) to 0.75 (+/- 0.23) in children with vestibular hypfunction. Analyzing Cronbach's Alpha, a 99.6% (α = 0.996) true score variance was achieved when two impulses were performed to the right and 98,1% to the left in healthy children and 99.9% in children with impaired vestibular function. CONCLUSION: These results indicate that two high velocity artifact-free impulses from an experienced tester are sufficient to evaluate vestibular function in difficult vHIT testing conditions like testing very young children. Further impulses do not improve test reliability.