Comparison of the video head impulse test with the caloric test in patients with sub-acute and chronic vestibular disorders.

The aim of this prospective register-based study was to compare video Head Impulse Tests (vHIT) with caloric tests on 173 patients assessed by a tertiary Neurology referral centre who had been referred for investigation of dizziness or vertigo and whose symptom duration was one month or longer. Abnormal vHIT was defined as angular velocity gain (peak eye velocity/peak head velocity) less than 0.79 at 80 ms and 0.75 at 60 ms, which was two standard deviations below our institutions' lower limit of normal; together with refixation saccades. Abnormal bi-thermal caloric testing defined unilateral hypofunction as a 25% difference using Jongkee's formula and bilateral hypofunction was defined by the sum of the peak slow phase velocities over the four irrigations being <20°/s. Sixty patients had abnormal results on one or both tests, of whom 51 had unilateral and nine bilateral hypofunction. With caloric testing considered as the gold standard, the sensitivity (95% CI) of the vHIT was 18/52, 34.6% (22.0-49.1), and the specificity (95% CI) was 113/121, 93.4% (87.4-97.1). However vHIT was more sensitive in the nine patients with bilateral hypofunction with 100% abnormal vHIT results while only 4/9, 44% had abnormal caloric results. In conclusion these results support the continued use of both vHIT and caloric tests in patients with sub-acute and chronic vestibular symptoms, especially if the vHIT is normal.

Are there diurnal variations in the results of the Dix-Hallpike manoeuvre?

This study examines diurnal variation in the result of the Dix-Hallpike manoeuvre when testing for benign paroxysmal positional vertigo in a randomised crossover study of the order of diagnostic testing in a community out-reach clinic for a tertiary neurological centre in Wellington, New Zealand. Study participants were adults referred for physiotherapy treatment. Dix-Hallpike manoeuvres were performed to both ears, and groups were randomly allocated to have a morning, then afternoon, sequence of testing or vice versa. The results of the Dix-Hallpike manoeuvres were digitally recorded and reviewed by a second blinded assessor. A total of 27 of 50 participants (54%) tested positive on at least one of the days, six of 27 (22%) had discordant results. The difference in marginal proportions was 0% (95% confidence interval: -9.6 to 9.6), p=1.0. The time of day is not a factor in false negative Dix-Hallpike manoeuvres, although 22% of those with positive results tested negative on one of the two measurements.