JC: Can I safely discharge dizzy patients from the ED?

dizziness stemlyns


Last month, I came across a clinical review published in the Annals of Emergency Medicine1 that got me really excited!

It got me excited (in a geeky professional sense of course…) as it covered a topic that causes significant diagnostic difficulties in the ED for emergency practitioners, senior and junior alike.

This topic is that of distinguishing between peripheral and central causes of vertigo (dizziness) in patients presenting to EDs with acute onset of symptoms. This distinction causes great anxiety amongst practitioners as it is notoriously difficult to differentiate between the two causes and the prospect of discharging a patient home with a posterior circulation stroke is quite frankly a scary one.

I would suggest however that the bedside diagnosis is actually relatively easy. I have come across many papers/reviews/YouTube videos and attended multiple conferences where this was covered but this is probably the best review I have read in recent years. I therefore propose to summarise it in this short blog post.

As always, I would also suggest you read the original paper which can be accessed here for sake of completeness.


  • three percent of ED patients present with dizziness, lightheadedness or imbalance
  • for patients presenting in the ED with acute dizziness, the crucial differential diagnosis for the emergency physician is vestibular neuritis versus posterior circulation stroke
  • misdiagnosis of posterior circulation stroke is a needle in haystack phenomenon (haystack = dizziness as presenting complaint, needle = posterior circulation stroke as a diagnosis)
  • 28 – 59% (!) of cerebellar strokes are misdiagnosed in the ED
  • posterior circulation strokes are missed more than twice as often as anterior circulation strokes
  • the personal, economic and public health burdens of misdiagnosis are huge


  • the words dizziness, lightheadedness or imbalance are not diagnostically meaningful
  • these words are used interchangeably by both patients and clinicians adding to the confusion and diagnostic difficulties
  • not all patients with acute onset of dizziness have nystagmus (regardless of the cause for dizziness)
  • various medical and neurologic conditions can cause an isolated vestibular syndrome (Wernicke encephalopathy without confusion, anticonvulsant toxicity, multiple sclerosis etc.)

Do you see yet how potentially complex this could be?  At the same time, the author suggests in his review that physical examination outperforms MRI scan for diagnosis in the first 48 hours!

And now do you see why I got so excited coming across this review? 🙂


The author simply suggests to use the below screening tool which consists of five questions to which you can get a dichotomous answer only:

  • is there a central pattern of nystagmus?
  • is skew deviation present?
  • is the head impulse test negative (in patients with nystagmus)?
  • are there any CNS signs on focused neurological exam?
  • can the patient sit and walk unaided?

This is reproduced in a very good flowchart in the original article I strongly suggest you consult.

Do not get too worried if you do not understand it fully yet. Just keep reading…

As explained earlier in this post, it is worth mentioning that no single element of the history distinguishes reliably neuritis from stroke so you can probably forget all the traditional teaching given to you on how to do this based on history, risk factors, severity of symptoms  etc. It is simply too unreliable.

The author however suggests that physical examination can accurately distinguish between the two entities and this is largely based on the HINTS test (Head Impulse Test, Nystagmus and Test of Skew). FOAMCast has a great summary here which links to other FOAm resources too.

The presence or absence of nystagmus is key as the HINTS test can only be only performed in patients WITH nystagmus. The author suggests to have additional tests at the bottom of the flowchart for those patients WITHOUT nystagmus.

Again, keep reading (I hope you’re still following me)…

The author mentions here (and I agree with him) that the elements of the HINTS test are not traditionally taught to EPs and (so far) only stroke neurologists could accurately distinguish the two entities using this test. I have myself experienced a huge knowledge gap in juniors and seniors alike when it comes to using this test in the ED. Just like the author, I would however argue that it is a fixable knowledge gap and we can learn to understand, perform and interpret the results of this mystical test.

After all, it is NOT the stroke neurologist who will seeing the acutely dizzy patient in the ED but indeed the emergency physician. It is therefore your duty to become acquainted with the below-described diagnostic technique if you do not want to misdiagnose your patients.

Breaking down for you the flowchart into simple questions/steps:

Step 1:  is there a central pattern of nystagmus?

Despite the HINTS acronym, the author suggests you start with this assessment and this has been my clinical practice too. The reason behind this is this is probably the easiest and less intrusive test to carry out. The other reason is (as mentioned above) that you can carry out a HINTS test in patient with nystagmus only. So if there is no nystagmus, then you should straight skip to step 4 and 5.

If the answer however is yes to step 1, the cause is presumed to be central and the patient must be assumed to have a posterior circulation stroke.

The next question I hear you ask is: how can I differentiate between central and peripheral causes of nystagmus?

Very broadly speaking, the unidirectional gaze-evoked horizontal one is the only one that should re-assure you (as it points to a peripheral cause). All the others (direction changing, rotational, vertical etc) should start worrying you.

This is how I remember this but it is a very crude representation of something a bit more complex: I strongly suggest you consult and refer to the table found in the original article.


no nystagmus normal finding rules out a vestibular neuritis but consistent with cerebellar stroke
spontaneous horizontal nystagmus in primary gaze does not distinguish more commonly observed with peripheral causes but is not diagnostic
gaze-evoked horizontal nystagmus that beats in only one direction does not distinguish suggest a peripheral cause but is not diagnostic
direction-changing gaze evoked horizontal nystagmus central this is always central but can be benign in cases of intoxication (ethanol, anticonvulsant)
pure vertical nystagmus central central cause always
torsional (also called rotatory) nystagmus central expected finding in BPPV but these patients do NOT present but acute vestibular syndrome but a triggered episodic vestibular syndrome
skew deviation normally absent; its presence means a central cause not very sensitive but if present then means a central cause

Step 2: the skew deviation test

Again, this does not follow the order of the letters in the acronym but I do suggest you do this one next (this is the S so the last one in the HINTS). This is simple to do: it is an alternate cover test, looking for small vertical corrections of the target by the uncovered eye. This is is highly specific for a central cause (although not very sensitive) so if this is positive, you should continue to be worried about a central cause.

This YouTube video gives a nice demonstration of how to perform this test.

Step 3: the head impulse test

This is probably the most difficult to carry out (though you should be able to get better with practice) and this is the one that causes the most confusion as it is counterintuitive. It is a real headache to understand and remember for those who do not perform this test regularly (Ed – I always look this up for a refresher when assessing a patient with dizziness).

If I didn’t get your attention so far, I need you to focus now!

This is why it is counterintuitive:

  • a positive (abnormal test) suggests a peripheral cause. You should worry less as it is likely to be a vestibular neuritis.
  • a negative (normal) test should worry you as it suggests a central cause. This is because the reflex does not loop through the cerebellum and is therefore intact in a posterior circulation stroke.

This is how you carry out the test:

  1. have the patient fix their gaze on your nose
  2. hold their head and rapidly but minimally (this is important as your patient with dizziness will resist you) turn their head side-to-side by 10 – 15 degrees so they cannot predict the changing directions. I usually sit the patient on the edge of the examination couch (if they can) and warn them I have to get very close to them for this strange test (you need to be able to see their eyes).

The normal/negative is when the patient’s eyes stay locked on target (which is your big nose) and the positive/peripheral is when the eyes move with the head and there is a correctional saccade to bring the gaze back to your nose.

I have highlighted some letters on the above paragraph as it is how I used to remember what the results meant when I started doing this. Please ignore if you have your own way to remember this rather counter-intuitive test.

Stick to abnormal test is good, normal test is bad (even writing this down does not feel right!) 🙂

The saccade should be easily observable but there are suggestions from our FOAMed friends online that you could use the Slo-Mo function of your iPhone camera if you struggle with this. Scott Weingart made a great video of how to use the iPhone to record this (you will need a second pair of hands to perform the test or do the recording)

You really do need to understand, practice and master this test (probably in this order). It is important to explain what you are going to do to the patient as they will naturally resist you because of the very nature of their symptoms and most of your patients will be elderly with limited neck movements – be gentle!

Step 4: the focused neurological exam for patients without nystagmus

We mentioned above that the HINTS test can only be accurately carried out in patients with nystagmus. In those patients who present to you without nystagmus, you should skip step 1-3 and do a focused exam. Here you will specially test the cranial nerves for hearing changes, anisocoria, altered phonation and facial loss of pain or temperature sensation, plus cerebellar ataxia.

Step 5: test for ataxia and gait

Is the patient able to sit up unaided or walk unassisted? I usually simply ask the patient to sit up on the examination couch after reassurance and if this works, I test their gait. These can be challenging as they can be extremely dizzy in presentation and not very keen to cooperate with you. Patients with acute neuritis can usually walk though.

Regardless of this, however, a severe ataxia and/or gait inability would be a good reason to admit not only for safety reasons but further investigations too as you would not able to completely rule out a stroke.


I hope that I have managed to summarise a topic that covers an area of emergency medicine that still causes great difficulty to understand for EPs of all grades. It is important that you familiarise yourselves with the above diagnostic techniques as you will be seeing these patients presenting to your ED.

This is something you cannot just read and assume you will be able to perform at the bedside (even if you understand it). You will need to practice, practice and practice again… And refer to this post whenever you need to.



@baombejp on Twitter

Before you go please don’t forget to…

Further reading



Edlow JA. Diagnosing Patients With Acute-Onset Persistent Dizziness. A. November 2017. doi: 10.1016/j.annemergmed.2017.10.012

Cite this article as: Janos Baombe, "JC: Can I safely discharge dizzy patients from the ED?," in St.Emlyn's, January 12, 2018, https://www.stemlynsblog.org/dizziness/.

6 thoughts on “JC: Can I safely discharge dizzy patients from the ED?”

  1. Pingback: Yrsel – Översikt – Mind palace of an ER doc

  2. Peter Tagmose Thomsen

    Thanks for the this!!

    I’ve worked in the neurological emergency department, and the (true) vertigo patients were often the hardest to rule out and sent home
    The Kattah et al paper of 2009 that proposed that the HINTS was superior to an MRI, used a neuro-ophtalmology for the HINTS – which does raise the question of generalizability. It all usually comes down to whether you catch the positive HIT during your exam on the unilateral nystagmus patient, or you do it wrong / patient can’t relax / you don’t see the lagging eyemovement. I always wondered why we didn’t use the neuro-ENT’s special googles for detecting the subtle HIT positive tests . If the HIT is negative (= stroke is possible), then we usually would do the scan depending on history og neurological findings.
    We would usually also use the TiTrATE method: https://www.ncbi.nlm.nih.gov/pubmed/26231273

    I cannot access the paper of 2017 you refer to at this time (sorry) – do they have any reason to exclude the patients without nystagmus?

    In Kattah et al they describe many of their patients to be neurologically without findings (ie no nystagmus, no ataxia):

    “Frontline misdiagnosis of posterior circulation strokes presenting with dizziness appears common, occurring in perhaps 35% of cases. The high rate of misdiagnosis may not be surprising given that 58% of patients in our series either had no obvious signs or had only isolated, severe truncal ataxia”

    “All patients were unsteady (i.e., broad-based gait or difficulty with tandem walking), but severe truncal ataxia (inability to sit without the use of arms or assistance) was seen only among those with central lesions ”

    So the question is – how many posterior stroke patients present with subtle easily missable / no neurological findings, and negative nystagmus test. As we did not know this number, we often scanned them (especially if high pre-test risk)

    Thanks again
    Peter, Denmark

    1. Hi Peter

      Thanks a lot for you comment.

      There is no doubt that there is a massive knowledge (and therefore skills) gap amongst emergency physicians in diagnosing posterior circulation stroke. This is because this test has been traditionally the remit of the neuro-stroke specialist and (as you mention) signs can be very subtle.
      In most, emergency departments in Europe, America, Australasia there is no direct access to a specialist 24/7 and these patients are seen by emergency physicians/general practitioners.
      My argument is it is therefore our duty to master this bedside test deemed mystic and unachievable by anyone else. The same way we mastered rapid sequence intubation or bedside sonography. These were deemed to be the remit of the anaesthetist or radiologist only but are now core components of our daily practice.
      We can achieve this only by spreading knowledge (hence my blog summarising the article), practice (on patients and at home). so I do believe we can overcome generabilasibility issues with time and practice…
      There are “tricks” to pick up the subtle saccades you mention. Like using an iPhone SLO-MO camera (see https://emcrit.org/racc/iphone-hit/) or simply by practicing the skill until one masters it.
      There is no doubt that just like in the case of ultrasound, this will be operator dependent but “practice makes perfect”. 🙂
      You can access the full article here http://www.annemergmed.com/article/S0196-0644(17)31795-X/fulltext for sake of completeness.

      Thanks again for your mcomment

  3. Pingback: Vértigo periférico vs central – Tiempo es Cerebro

  4. Pingback: St Emlyn’s January 2018 in review - St.Emlyn's

  5. Pingback: emDOCs.net – Emergency Medicine EducationEM Beats by Doge: Vertigo - Central or Peripheral - emDOCs.net - Emergency Medicine Education

Thanks so much for following. Viva la #FOAMed

Scroll to Top