FOAM4GP Map – Do dizzy patients make your head spin?

image credit: Jun07DSC_0212 via free images (license)
image credit: Jun07DSC_0212 via free images (license)

Patients presenting to General Practice with dizziness are a significant challenge. Dizziness incorporates a wide range of possible causes varying from completely benign to instantaneously life threatening. Appropriate workup involves comprehensive history taking and examination skills followed by judicious use of appropriate investigations.

While a large proportion of patients are eagerly awaiting their MRI form to rule out the ‘obvious’ brain cancer they suddenly developed at 3pm that afternoon (you are seeing the patient at 3:30pm), we are all better doctors with more tools available to us (including the grey matter between our ears) than to have to rely on expensive tests to make (sometimes) simple diagnoses.

Ok. So what tools do we use? What is the evidence for them? And how good are we at it??

Lady_Face_clip_art_smallSandra is a 55yo F who presents to your clinic at 3:30pm significantly upset and anxious. She tries to walk down to your room but keeps gently leaning into the wall on the left. She sits in the chair and looks significantly pale. “I’m so dizzy doctor! Make it stop!!”

1. So Sandra has presented to my clinic with dizziness. How likely is it to be something serious?

The three most common causes for dizziness presenting in Australian General Practice (accounting for 93% of presentations) includes: Benign Paroxysmal Positional Vertigo (BPPV), Vestibular Neuronitis, and Meniere’s disease (AFP Vertigo Part 1 2010). In the US with 2.6 million presentations to the emergency department (different cohort – higher pretest probability), 4% were diagnosed with a cerebrovascular accident (Emcrit). As for data on patients presenting to Australian General Practice with dizziness who are then diagnosed with a CVA, this is unknown!

Ok. So based on statistics all we can say is Sandra is likely to have a less than 7% risk of her dizziness being anything serious before we do anything (though we haven’t looked at risk factors etc yet).

2. So lets ask Sandra some history and see what can help.

Ok. So before I even ask a question, Sandra is >50yo. This must mean she is more likely to have a stroke or brain tumour right? Yes – ish.

Age? – While increasing age is associated with increasing risk of CVA, 25% of patients (15/69) in the largest available prospective trial of patients with ischaemic stroke were under the age of 50! (Emcrit 2011)

Right then. We need to work out if this is vertigo, presyncope, or disequilibrium. Don’t we? Do I ask about postural symptoms, room spinning, floor moving etc. Is it helpful?

Type? – Possibly not. Sadly, while this has been discussed in many many textbooks and is a regular teaching point, it is not evidence based. “[Type of dizziness has] been shown to be a very inaccurate measure and not a trustworthy predictor of underlying cause”. In a trial of 300 patients presenting to an ED in the US, approximately 50% were unable to give a history to discern between the different types (Emcrit 2011). Additionally, an article in J Gen Int Med 2008 suggested that differentiating a presentation into vertigo and then excluding cardiac causes as a consideration is fraught with risk and may miss significant cardiac causes of dizziness. A Mayo clinic study 2007 on imprecision of patient reports on dizziness symptom quality also stated that patient descriptions were often “unclear, inconsistent, and unreliable – they suggest focusing on timing and triggers over type” – Mayo clinic 2007 – Link

What about for working out presyncope vs vertigo to differentiate cardiovascular causes from other. Nope. Not alot of help here either…

A systematic review of ~1500 patients suggested that in patients with primary cardiovascular disorders (MI, orthostatic hypotension, or syncope) reported vertigo (yes vertigo… not presycope) in 63% of presentations; and in 37% of these patients, vertigo was the ONLY type of dizziness reported. This was from a Systematic Review in J Gen Intern Med 2008 (though arguably they were piecing together fairly diverse research to try to pull a conclusion. Most studies looking at syncope/presyncope actually excluded patients with vertiginous symptoms).

Ok honesty moment. Though the evidence doesn’t stack up I am stuck in my old ways. If a patient describes their sensation as they are seeing spots in vision, feel hot and flushed, and dizzy – I will probably be considering a presyncopal consideration. If the person is not dizzy but just falling over, I still will think things like peripheral neuropathy etc. But! I will always keep in mind now that my ‘typing’ can be wrong and therefore remain alert to red flags/central causes.


3. So if we are not working the dizziness into different types – what else will we structure our history around?

Central causes vs Peripheral causes.

This is a really important distinction to make and a major aim of our ‘triaging’. Central causes are generally the more concerning and more emergent, peripheral causes often give us a little more time to make the diagnosis.

Central causes – These are problems with brain stem, cerebellum etc. These are lesions which are generally brain problems and, subsequently, generally bad. Examples include: CVA/TIA, MS, brain tumours, vestibular migraine etc.

Peripheral causes – These are problems which affect the vestibular nerve and vestibular canals. These are generally more benign causes. Examples include: Labyrinthitis/vestibular neuronitis, menieres, BPPV etc.

Here is a great summary of all the different central and peripheral causes of dizziness. “The Ochsner Journal” 2009 – Link. I won’t be going through all the causes in this Foam4gp map.

If want to cut directly to the chase, here is a table of differentiating features between central and peripheral causes of dizziness. Evidence based – Link

Do you like summaries in table form for (almost) all of the evidence discussed here? Here is the link!

Lady_Face_clip_art_smallSandra sits down in the chair. She looks flushed and upset. She holds tightly to the side of my desk and makes a flippant comment about “not being like this since she was 18 and may have imbibed too much of something”. After a shared chuckle, she reports that ‘she wants to get off this ride’.

Sandra says her dizziness started this afternoon. It came on fairly quickly over 30-60 mins. She has felt mildly nauseous with this and feels the room is spinning around her. She says she also can’t stop sweating. Oh. And she feels like she has a weird buzzing in her ear. She reports she has had this once before ‘a few years ago’ which went away after a few hours.

So poor dizzy Sandra may have Meniere’s (spoiler warning… too late sorry). BUT can I happily say it definitely is, give her some stemetil, and send her on her merry way? (It is time for coffee anyway). Or will I awake in the middle of the night myself slightly dizzy and sweaty that I could have missed something TIA ish?


4. Anything else in the history we might check before we launch into further work up?

Firstly, always important to consider unexpected causes and our extra resources (previous notes). No head injuries, new medications, OTC medications, exposures to toxins and poisons, PMHx of previous TIA’s, blood thinners, neuroleptic medications etc. Always important to remember how helpful our notes are in General Practice. Can save a lot of fuss if there is an obvious cause or significant risk factor.

Sandra is a remarkably well woman. No regular medications, allergies, exposures, OTC or recreational drug use, minimal alcohol, doesn’t smoke the wacky tobaccie etc….

5. So from our bit of history. Can we start to work out Central vs Peripheral causes? What can we use?

Associated symptoms? Any FAST symptoms? Focal neuro symptoms? – While neurological signs ARE helpful to increase likelihood of a CVA, their absence is unhelpful.  – (Emcrit 2011).

Symptoms out of proportion? What if she won’t stop vomiting despite me asking nicely? – Some case studies have reported that the degree of dizziness, vomiting, or gait disturbance, out of proportion to degree of dizziness or nystagmus, may be a marker for cerebellar and central pathology – (Emcrit 2011)

Triggers? Does moving her head around make things worse?- These can be useful – for example for BPPV (head movement) and presyncope (rapidly standing). However, triggers are more non-specific when talking about vestibular neuronitis versus CVA. (Emcrit 2011)

Concurrent illness and exposures? Common sense type stuff. – There is no significant evidence to help us here but applying the ‘common sense blow torch’ is helpful (eg. recent ear surgery, now ear pain and dizziness, check their ear!) (Emcrit 2011)

Onset? Sandra’s dizziness came on slowly over time. It can’t be a TIA/CVA then. Can it? Poor evidence generally available but two retrospective studies suggested some link between abrupt onset dizziness and CVA cause. (Emcrit 2011)

Prodrome? She didn’t mention a recent viral illness so not vestibular neuronitis? – Not necessarily true. Patients have a 25% prevalence of a prodrome in vestibular neuronitis and similar prevalence prior to stroke interestingly (presumed TIA?)! (Emcrit 2011)

Auditory symptoms? Surely this lady just has Meniere’s… Why are we still talking about this?? – While historically auditory symptoms point to a peripheral cause, more recent research focusing on blood supply to the ear has decreased our reliance on this ‘differentiating’ sign – (Emcrit 2011)

Duration? This HAS been shown to have some diagnostic usefulness – more towards an actual diagnosis than central vs peripheral. Here is a suggested guide from BPAC(NZ)

Seconds (psychogenic), Less than 1 min (BPPV), Minutes (Vascular/Ischaemic), Hours (Meniere’s or vestibular migraine), Hours to days (vestibular neuronitis, central causes, MS), Recurrent with headaches (Vestibular migraines).

So what else can we use that is evidence based to make our diagnosis? Here are some summaries from a great review of the evidence in Emcrit 2011.



Here is a history taking algorithm from AFP for those more visually inclined…

Screen Shot 2015-05-10 at 9.01.46 am

6. Great. Cheers Rob. So now I have no idea what to use? So what CAN we do then?

While a lot of the research talks about UNhelpful history questions, I would suggest that like everything in medicine we are trying to create a clinical picture to fit a diagnosis. Sandra presents with onset over 1hr (suggests meniere’s), auditory tinnitus (suggests meniere’s), no neurological/presyncopal/preceeding symptoms, and no other red flags. So while in isolation these things may be questionable – together they may add up to help us suspect a diagnosis.

Lady_Face_clip_art_small7. Sandra is now clutching a vomit bag (as you have seen where the outcome of your upcoming exam can end up – usually the floor.) So we would like to examine Sandra. We move her over to our couch very slowly and collect a few odd looking things.  Sandra looks at us very strangely – hat pins, tuning forks, otoscopes, stethoscope, cotton buds, sharp/dull points etc. Sandra helpfully asks whether you were planning to examine her. Or torture her…

In reality, this is where the evidence lies for helping us to discern central vs peripheral causes. So lets dive right in.

Imbalance and gait disturbance. Sandra wobbles over to the couch (not one particular way) – Severe truncal ataxia (inability to sit with arms crossed unaided) was the only ‘obvious’ sign in 29% of pts with stroke and more frequent than in patients with vestibular neuronitis (33% vs 0% p < 0.001). Patients presenting with dizziness and significant imbalance and gait unsteadiness was associated with stroke (OR 3.71, 95% CI 1.30–10.65). (Emcrit 2011)

Neurological signs. You put Sandra through your usual battery of CNS/Upper limb/Lower limb (including gait, balance, co-ordination, and my favorite word in medicine, dysdiadochokinesis). She asks if you are opening a gym soon… – In the systematic review (Emcrit 2011), focal neurological signs were reported in 80% of CVA patients (though significant diagnostic bias needs to be considered – eg. patients and doctors, after they have been told the patient has had a stroke, are more likely to attribute ‘soft’ neurological signs to actual pathology). “Symptoms and signs of CVA in the posterior circulation are generally under-appreciated by patients, physicians, and even standard rating scales” – I will discuss these further below so you are not one of these!

Internuclear Ophthalmoplegia – This is a specific sign for MS. Here is a link to a refresher video of this clinical sign – Link

Nystagmus details: – There was insufficient evidence to suggest the vector, pattern, and fixation characteristics of nystagmus, either help or hinder diagnosis of central vs peripheral causes. Hopefully more research will help clarify this. (See “Gaze evoked nystagmus” as a special case below). Nystagmus refresher here in this free to access power point presentation – University of Iowa 2015

Otoscopy: Can be useful if you see vesicles (Ramsay-Hunt syndrome), Cholesteatoma, or Otosclerosis.

Audiovestibular testing: Do we send them off for a formal hearing assessment? – No specific evidence to routinely recommend this for evaluation of the patient with dizziness

Ok so time to pay attention everyone. Everyone is talking about the HINTS assessment. What is it?

Here is a brilliant summary from LITFL and also EM Nerd

What? – It is a set of three tests called “Head Impulse, Nystagmus, and Test of Skew.” It has been claimed to have better sensitivity and specificity for picking up stroke than an MRI. (Ah but everything has a catch…)

Who? – Patients presenting to ED with what they defined as “Acute Vestibular Syndrome (vertigo, nystagmus, nausea/vomiting, head-motion intolerance, unsteady gait)” and >1 risk factor for stroke (these are discussed in the article below).

Why? – Because you are likely to hear more and more about these set of tests. How good is the evidence? Listen to a critique of all evidence in this great podcast here – EMJournal club

How? – Aha! Look below…

Vestibulo-ocular reflex: – “This was the best bedside predictor of peripheral versus central causes of acute vestibular syndrome”. The head thrust test had a sensitivity for CVA approaching that of an MRI and both having similar specificity!! Normal response helps confirm a stroke and abnormal response suggests a peripheral lesion.

How do you do this? Check out this video:

This great website with free neurology info is great for refreshing the neurology of the vestibulo-ocular reflex (I know I had to look it up again!) – From Teach Neurology – great online FOAM blog – Link

Gaze evoked nystagmus: – (More useful) – Direction changing nystagmus on lateral gaze has been shown to be highly specific (92%) but not very sensitive (38%) for central causes. Great description of the neurophysiology here – Teach Neurology – Link

Skew Deviation: This results from a right-left imbalance in inputs to the oculomotor system and, with very rare exceptions, is central in origin. Its specificity is 98% but a sensitivity of only 30%. Have a look at the video for explanation and again the Teach Neurology link for neurophysiology.

Hallpike testing: A useful test for diagnosing BPPV but not helpful for differentiating between central and peripheral causes. Surprisingly, while many documents report that this has been the ‘gold standard’ for diagnosing BPPV for many years, actually finding some evidence for its use was very difficult! It has been reported that it has a sensitivity of 82% and specificity of 71% but that this is highly operator variable – Otolaryngology 2008. Factors include: speed of movement, time of day (not sure why!?), and angle of the plane of the occiput during movements.

Some people have suggested using a response to Epley after diagnosis of BPPV, to help confirm the diagnosis. I cannot find any current evidence on the Epley as a diagnostic tool – more as a therapeutic one.

Final word on HINTS: While the evidence is strong and the tests are available bedside tests – I am not sure that I would be relying on my interpretation of the results. The reason is that the trials had two factors different to my clinic work. Firstly, they were ED presentations (and therefore higher likelihood of CVA) and also the tests were performed by 1. ED doctors who had had 4 hrs training 2. Specialists in neuro-opthalmology. I would be a little concerned that my interpretation of the youtube videos may not directly correlate. So while a positive may be helpful, a negative (for me) may not.

Cardiac examination:

We all routinely do one – is there any evidence for it?

Most resources I have seen suggest at least a Heart rate and rhythm (+/- ECG (see below)), BP standing and supine (3 mins each position – that is a long time in a consult? Practice nurse help maybe?) and auscultation on the neck (they suggest for ?TIA/Stroke risk factor if bruit) – BPAC NZ Dizziness 2012

The above article did not suggest heart auscultation. If you are considering a possible cardiac cause and we cannot rely on typing our dizziness (as discussed above) into ruling out syncope – it would also seem a prudent and quick examination to perform? (Happy to be corrected if people have a different view – I cannot find any specific evidence on this?!)

Should we ECG everyone? A critical review by Southern Medical Journal 2000 suggests that because cardiac arrhythmias are rarely the cause of dizziness (1-5%) of patients, that routine ECG +/- holter ‘seems less important in working up the patient with dizziness without syncope that in the evaluation of the true syncope’. Truth moment – I find ECG a quick and easy thing to get where I am so would probably consider it more often that not – low chance of harm. Though one patient asked me once whether I was going to electrocute them after putting the dots on….

Carotid bruits? No evidence either way but as a marker of atherosclerotic disease seems a prudent thing to do. Specificity/sensitivity for symptomatic bruit is approx 75%/75% – BMJ Practical Neurology 2002

Here is a great article that discusses the possible work up of a cardiac cause for dizziness – with a focus on presyncopal and syncopal symptoms (though above stated this is hard to define – I still think it is a useful algorithm to follow if suspect red flags) – RACGP AFP 2014 Dizziness

8. Want to do all of the above? Apparently it can be done and all in a ‘Jamie Oliver 15 mins meal’ kinda way… Here is the summary for you rapid people out there…

Here is a link to an article on Medscape explaining how to do it – Link  (I’d actually be more impressed if someone can actually do a Jamie Oliver 15 mins meal in 15 mins. My wife includes the time for cleaning up the bomb site which is my cooking canvas…)

9. So what about some blood tests? Patients love blood tests don’t they?

In one meta-analysis, only 26 of 4,538 patients (0.6 percent) had laboratory abnormalities that explained their dizziness (AAFP 2010 – Link)

10. Ok. So now I’m getting confused by all this talk of different tests etc. I’m just going to send her for a CT….

Hold on!!

CT Scanning – CT scans have low sensitivity (approximately 16%) for acute infarction,7 particularly in the posterior fossa (Edlow JA, Newman-Toker DE, Savitz SI. Diagnosis and initial management of cerebellar infarction. Lancet Neurol. 2008; 7: 951–964). So this is not going to help you if you are thinking about a possible TIA/CVA (though these are probably people you are sending to ED in the real world).

10b. Could we order an MRI (which is not on the MBS – by the way) to make us feel better – and if negative then reassure the patient and sleep better? (The patient of course…)

MRI Scanning – MRI has a sensitivity of 83% for CVA (particularly finding ischaemic areas in the posterior fossa). Further, MRI performed within the first 48hrs after onset of dizziness missed a CVA in 8/69 patients who had a follow up MRI. Interestingly, taking this into account, the MRI was WORSE at ruling OUT a stroke than the composite HINTS evaluation! (Emcrit 2011)

11. What about some treatment advice depending on what we have found?

Here are a collection of links to treatment options for some of the most common conditions:

Meniere’s –

Vestibular Neuritis –


Posterior circulation stroke – BMJ

Vestibular Migraine –

eTG – Very good for all of the above if you have access (not free).

12. Summary of everything

Here is a great summary from the Royal Victorian eye and ear hospital and information about referral planning – Link

Lady_Face_clip_art_smallSo Sandra has patiently allowed you to perform all of the above examination and bedside tests. All have been negative and her dizziness has continued intermittently associated with the constant ringing in her ears. I sit her down and gently talk about the possibility of Meniere’s, explain the basic pathophysiology using diagrams of the ear, print her an information sheet on the same, talk about simple measures she can try including white noise etc. for her tinnitus. Discussed importance of following her up if this is an ongoing problem (to monitor her mood given significant psychological impact of this illness), and finally to offer referral to ENT if ongoing trouble and she would consider surgical options/other treatments. You ensure she is safe, discuss importance of having someone else drive for her while she is having these attacks and sensibly if she feels any of these symptoms developing at any time that she pulls off the road until symptoms settle for safety.

2008-03-20Rob’s final word…

So dizziness is a very difficult presenting symptom. I have read through a wide array of evidence and guidelines and I have mixed feelings about my certainty in making an evidence based diagnosis. I was disappointed that the evidence for ‘typing’ dizziness has a weak evidence base as I was taught this from medical school.

I feel that by performing an appropriate history and examination and always considering red flag features or ‘things that don’t fit’ then I will do my best as a doctor to discern the serious from the benign. My primary aim in these patients will be to discern central from peripheral causes. If I have suspicions of possible central causes I will send to the local Emergency Department with a suggestion for an MRI.

I think dizziness is a high level experiential GP skill where seeing patterns of disease may be helpful. I also suspect, that further research in the future will help us highlight the best ways to make more certain diagnoses. It is disappointing that there is not more GP-centric (and perhaps Australian) information about dizziness presentations as this is such a common presentation and such research would be very helpful. Most of our information currently is extrapolated from Emergency Department presentations which are a different cohort.

Hope this has been of interest to everyone. Please leave lots of comments – discussion about dizziness. Anecdotes are always useful too. I would argue this is one of the most difficult presentations? What do you think?

Til next time,

Cheers, Rob.


5 thoughts on “FOAM4GP Map – Do dizzy patients make your head spin?

  1. Another utterly brilliant post on a much needed topic, Rob, thank you. As usual, the perfect combination of evidence based medicine, helpful links, clinical relevance, personal experience and humour! I’ll certainly be recommending it to our registrars.

  2. Wish I’d had this as a student! But now, feel my practice will be much improved for thinking all this thru with your guide. The links are great too! Ta! 🌻

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s