Quantitative vestibular assessment: The development and validation of a novel, remote video head impulse test against in-clinic measurements.

Abstract

To develop a novel remote head impulse test (rHIT), and to provide preliminary data validating the rHIT vestibular-ocular reflex (VOR) gains against the in-clinic vHIT. A convenience sample of 10 patients referred for vestibular assessment at our institution was recruited. In-clinic vHIT was used to quantify lateral VOR gains. Patients subsequently underwent an rHIT protocol, whereby patients performed active, lateral head rotations while their eyes and heads were recorded using a laptop camera and video-conferencing software. The vHIT and rHIT VOR gains were compared using paired t-tests, and a Pearson correlation coefficient between the gains was calculated. Absolute accuracy, sensitivity, and specificity of the rHIT were additionally calculated. Of the 10 patients recruited, 4 were male, and the average ± standard deviation (SD) age was 61.4 ± 15.3 years. As determined by the vHIT, 2 patients had normal bilateral VOR gains, 6 with unilateral vestibular hypofunction, and 2 with bilateral vestibular hypofunction. The correlation between the rHIT and vHIT gains was 0.73 (p < .001). The rHIT exhibited an absolute accuracy of 75.0%, sensitivity of 70.0%, and specificity of 80.0%. When ears had a vHIT VOR gain less than 0.40, the rHIT exhibited 100.0% accuracy. Conversely, 60.0% of deficient ears with vHIT VOR gains greater than 0.40 were incorrectly categorized by the rHIT. The rHIT may be better suited for detecting more severe vestibular deficiencies. Future iterations of the rHIT should aim to increase the video frame-rate capabilities to detect subtler VOR impairments. 4.