Role of the Patient's History of Vestibular Symptoms in the Clinical Evaluation of the Bedside Head-Impulse Test.

OBJECTIVE: Our aim was to identify the role of the investigators' knowledge of the patient's history of vestibular symptoms (PVH) in the clinical evaluation of the bedside head-impulse test (bHIT). We hypothesized that this knowledge will reduce uncertainty and improve bHIT accuracy when compared to quantitative analysis of the vestibulo-ocular reflex by video head-impulse test (vHIT).

METHODS: We looked for changes in the clinical assessment of the bHIT in 594 consecutive patients before and after taking PVH. bHIT was performed by 12 clinical neurologists with various clinical experience in neuro-otological diseases (novices to long-standing experts). vHIT was analyzed by four experts being blinded for the patients' clinical presentation and history of symptoms. The confidence of bHIT and vHIT was rated (0-100%).

RESULTS: One hundred fifty-four (15%) of 1,030 bHIT of all eligible patients (n = 515) were rated pathological. Thirty-five (22.7%) of them were rated bilateral vestibulopathies. Sensitivity of bHIT reached 56.3%, its specificity 92.4%; the positive predictive value (PPV) was 41.5% and the negative predictive value 95.7%. These data did not differ between bHIT before and after PVH. bHIT after PVH (post-bHIT) differed from pre-bHIT in 44.3%, usually with regard to the level of confidence but also in polarity (5%). The accuracy of changes in bHIT depended on the direction of change: a "normal" post-bHIT was correct in 92.3% while only 39.8% of pathological post-bHIT were pathological on vHIT. However, sensitivity of a pathological post-bHIT depended on the clinical experience in taking PVH and bHIT: the PPV was 20.5% in novices as compared to 69.6% in experts.

CONCLUSION: The study shows that PVH changes the certainty and/or polarity of the clinical evaluation of bHIT. Unlike expected, the increase in confidence in post-bHIT is associated with a consistently high specificity but no increase in sensitivity. Accuracy of changes in post-bHIT depends on the investigators' clinical experience: it increases only in experts but not novices. Since novices show only a poor PPV and moderate sensitivity of bHIT, pathological bHITs should be controlled by vHIT, even in patients with a positive PVH. By contrast, confirmed normal post-bHIT is usually correct.