Just Added!

A Comprehensive Guide to Resuscitation


Can We Detect Central Vertigo on Exam?

August 16, 2021

Written by Andy Hogan

Spoon Feed
HINTS and STANDING exams performed by emergency physicians (EPs) formally trained in both techniques demonstrated very high sensitivity for detecting central causes of vertigo in the ED.

Why does this matter?
Differentiating posterior circulation stroke from peripheral vertigo is challenging, and 10-30% of these strokes are missed in the ED. Studies have shown that EPs frequently misuse or misinterpret important techniques like the HINTS exam when evaluating patients with vertigo. This study suggests EPs, after completing focused training on the HINTS and STANDING exams, can reliably use these tests to rule out central vertigo in many cases (see STANDING Figure and tortured acrostic). 

Take a HINT…S exam training to improve rule out of posterior stroke
This single-center study analyzed 300 prospectively enrolled ED patients presenting with vertigo. All underwent HINTS and STANDING exams in the ED as well as an eventual confirmatory brain MRI. HINTS and STANDING exams were administered by study team members with 6 hours of dedicated training on both techniques.

The HINTS exam demonstrated sensitivity of 96.7% for central vertigo, which rose to 97.9% when non-stroke etiologies of central vertigo were excluded. This latter value corresponds to a NPV of 99.4%, suggesting that a patient with overtly peripheral findings on HINTS exam had <1% chance of having a posterior circulation stroke. The STANDING exam, which uses Frenzel glasses to enhance detection of nystagmus but does not test for skew, exhibited lower sensitivity (93.4%) but higher specificity (75%) than the HINTS exam (67.4%).

The single-center nature of this study limits its broad applicability. The findings do, however, suggest that formal training in the HINTS exam can significantly improve an EP’s ability to rule out posterior circulation stroke without MRI or a specialist consult. Adding Frenzel glasses to the ED toolkit may also improve detection of nystagmus and reduce false positive HINTS results.

Differentiating central from peripheral causes of acute vertigo in an emergency setting with the HINTS, STANDING, and ABCD2 tests: A diagnostic cohort study [published online ahead of print, 2021 Jul 10]. Acad Emerg Med. 2021;10.1111/acem.14337. doi:10.1111/acem.14337

3 thoughts on “Can We Detect Central Vertigo on Exam?

  • I completely agree that it’s training (and experience) that makes some neurologists very good at bedside assessment of vertigo, and that emergency physicians can certainly be trained as well.

    I also have no interest in seen a for profit vertigo certification process. However, I would like to see increased appropriate vertigo education at all levels of training for every emergency physician, neurologist and head and neck surgeon.

    A few points I would like to make about Gerlier’s study in question.

    1. HINTS was applied to patients who did not have spontaneous or gaze evoked nystagmus. The authors of the original HINTS studies have repeatedly stated this should not be done, as I describe in one of my videos. https://youtu.be/L4nOD8YdV-s This led to more cases of BPPV being in the Gerlier study than vestibular neuritis and central lesions combined. BPPV does not cause the Acute Vestibular Syndrome (AVS) with spontaneous or gaze evoked nystagmus, which is the only situation that the HINTS exam should be applied. So, as you pointed out, the HINTS exam in this study had a high sensitivity and very good NPV with regards to identifying central causes of vertigo. But it also had a poor specificity of only 67.4%. This was because of the high number of BPPV patients included in the study to which the HINTS exam should not be applied. And also somehow the majority of BPPV patients were found to be overall HINTS peripheral. But the only way to do that is to have an abnormal HIT, which is not seen in patients with BPPV. I emailed the author about this, and they had no explanation of how this occurred.
    2. STANDING was designed to diagnose BPPV as well as identify central causes. Vanni’s s STANDING study is unique, in that it is the only study ever conducted where they trained ED docs to diagnose both posterior canal (with the Dix-Hallpike test) and horizontal canal (with the supine roll test) as well as how to perform the Head Impulse Test. So in Gerlier’s study STANDING had only half as many false positive central results compared to HINTS when applied to BPPV patients.

    The other point I would like to make it that the first line of defense against missing a central cause of vertigo is NOT the HINTS exam, or STANDING. It is to screen all patients with vertigo for central features that would not be consistent with BPPV or vestibular neuritis and thus mandate a workup for stroke. These are listed in the central part of my algorithm which is in the current edition of Tintinalli, as well as in my CMAJ article, Figure 1. https://www.cmaj.ca/content/192/8/e182/tab-figures-data
    These central features are new significant headache or neck pain, unable to walk unaided, spontaneous vertical nystagmus, focal weakness or paresthesia, diplopia, dysarthria, dysmetria, dysphagia or dysphonia.

    In Gerlier’s study, they stated they excluded patients with localizing neurologic signs (which the majority dizzy strokes have) and patients who are not symptomatic at the time of examination (which most BPPV patients would be). So in most clinical settings, if you screened patient with these criteria, most would have vestibular neuritis, fewer would have non-localizing cerebellar strokes, very few would have BPPV. But her study had 90 patients with BPPV, 42 ischemic strokes, and only 26 vestibular neuritis/labyrinthitis. So something odd is happening here.

    I suspect that if all the patients were properly screened for central features as in my algorithm, and then HINTS was applied to only those with spontaneous or gaze evoked nystagmus, and positional testing was only applied to those with no vertigo or nystagmus at rest, the results would have been better.

    Peter Johns MD
    Department of Emergency Medicine
    University of Ottawa

What are your thoughts?