Written by Clay Smith
The HINTS exam alone, when assessed by emergency physicians in patients with acute vestibular syndromes (AVS), was not sufficient to rule out a central cause for vertigo. Have a low threshold to obtain MRI in patients with AVS and stroke risk factors.
Why does this matter?
The HINTS exam has been touted as an effective and accurate way to detect vertebrobasilar stroke vs a peripheral vertigo syndrome. See the Newman-Toker article. In the hands of a neuro-ophthalmologist, that may be true. How do emergency physicians do with HINTS?
HINT – be careful with this
This was a meta-analysis of 5 studies of patients with vertigo who had a HINTS exam as part of the workup and CT or MRI as the gold standard. For studies with only neurologists or neuro-ophthalmologists performing the HINTS exam, pooled sensitivity and specificity were 96.7 (93.1-98.5) and 94.8 (91-97.1), respectively. However, the only study which included both emergency physicians and neurologists had lower sensitivity and specificity: 83.3 (63.1-93.6) and 43.8 (36.7-51.2), respectively. Overall, quality of the studies was low, with all at risk of several biases: verification bias, spectrum bias, and detection bias, all of which would inflate sensitivity and specificity.
Can emergency physicians accurately rule out a central cause of vertigo using the HINTS exam? A systematic review and meta-analysis. Acad Emerg Med. 2020 Mar 13. doi: 10.1111/acem.13960. [Epub ahead of print]
Open in Read by QxMD
3 thoughts on “Does HINTS Miss Stroke?”
Thanks for bringing up the HINTS exam in vertigo. This study should not be used an excuse to not learn how to perform HINTS, but in fact, a call to action for more training in this and other bedside tests of vertigo.
Those who are old enough (me) can remember when no emergency physicians every handled an u/s probe. And to this day, many emergency physicians of my generation do not feel comfortable using u/s for many situations whereas recent grads u/s anything that moves.
There have been too many to count studies showing that emergency physicians over and under diagnose BPPV, as well as cerebellar stroke.
So we just throw up our hands and MRI everyone? Not a great strategy.
For decades all medical students and trainees in every specialty were taught myths about vertigo which has been ingrained in the assessment of vertigo leading to the deplorable situation that I described.
Now we have tools that are MUCH easier to learn than u/s that can diagnose and cure the most common cause of vertigo (BPPV) and rule out the most common of the dangerous causes (cerebellar stroke).
See my recent article in the CMAJ how to them.
And the equipment needed are your own two hands, your own two eye sand, perhaps the hardest equipment to persuade emergency physicians to use, their will.
The will to unlearn most of the myths they have learned for years and embrace new information that will completely change their view of seeing the vertigo patient.
Having taught how to perform HINTS to hundreds of people, from senior MD’s such as myself to medical students, I can tell you the same phenomena as with u/s will play out with vertigo. The young ones embrace it, the older ones generally, not so much.
There is a saying: "Science moves forward one funeral at a time"
I know when I retire in the next 5 years, there will be one less emergency MD who is crappy at doing u/s. But it my hope there will be tens of thousands of young emergency physicians who can properly diagnose BPPV and cerebellar stroke using their hands and eyes.
Thank you so much for this reply. I think you’re right. Personally, I know I need to apply myself to really learn to do this right. By the way, I feel your pain when it comes to ultrasound.
I’m not sure what to make of this article, this is clearly a difficult subject/diagnosis. The Newman article suggests MRI only had a sensitivity of about 76%. But this article used the MRI as the gold standard to test a tool that reportedly has a higher sensitivity? That methodology sounds kind of suspect. Even if HINTS has a sensitivity of 83%, that may be the best tool we have, especially if MRI really is only 76% sensitive. I really think more training in HINTS is needed and more studies are needed with very close attention paid to deciding on a reasonable gold standard.