Author: Andy Neill / Codes:CAP17, HAP17, NeuC2, NeuP2, PhC2, SLO1 / Published: 06/02/2018
This is a hugely common presentation in the ED and often one many doctors try to avoid. Many of us love the “juicy minor” (a great term, kudos to my former colleague Zeshan Ali) – the minor injury with a clear pathway of what to do. The patient fell hurt this bit of their limb and you do a pokey pokey at the sore bit and do an X-ray. You check to see the radiographer hasn’t put a red dot on it and give the patient the wonderful news that their appendage isn’t broken. They feel satisfied that they’ve had a doctory experience and more importantly you feel you’ve done a good doctory thing.
Contrast that with the headache. Most of them are miserable, so they’re grumpy and fed up and they’ve probably already taken some paracetamol and an NSAID so immediately your go to feel good options are gone…
Headache is also a presentation which is dominated by symptoms rather than signs. Patients will tell you all manner of things about their headache (in great detail if you ask the right questions) but rarely will you be able to demonstrate anything on exam that might make this an easy sell for a CT or an admission.
Remember this is one of our key fall back positions in EM when encountered with something difficult – we make it an SEP – Somebody Else’s Problem. Thus absolving the doctor involved of any responsibility to work out what’s actually going on… this is usually a turn to the dark side in EM so beware of it.
We also don’t have a nice easy test that we can bounce the decision to. While the CT may have become the ankle X-ray of headaches it actually isn’t much use in making a diagnosis for the vast majority of ambulatory headache patients.
So this is where we’re at. This is why the headache intimidates some of the new (and occasionally old) folk in the ED. We’ve a patient with miserable symptoms often resistant to our default paracetamol/NSAID. We have lots of symptoms and no signs and no clear way to make it an SEP. They’re expecting to feel better and they’re expecting an answer.
First we need a list of ideas. Before you go in the room have a list of potential diagnoses in your head and then question your way through them.
Common and benign
- tension headache
Common(ish) and serious
Uncommon and serious
- Pseudotumour cerebri (or whatever it’s called these days…)
- low CSF pressure headache
- Sheehan’s syndrome
- (consults wikipedia for increasingly obscure causes of headache…)
For lots of these, especially the uncommon ones, we won’t be able to make the final diagnosis in the ED. But remember that’s not what we have to do in ED – we need to pick up immediate life threats and know when to refer the others.
This is where ‘red flags’ come in.
Red flags first became a thing (I think) for low back pain. Components in the history that suggest the possibility of more serious disease. 90% of people will have an episode of low back pain in their life. The vast majority are completely benign. But there are some patterns of presentation that should ring alarm bells and we call these red flags. For example in someone with severe back pain and urinary incontinence we worry about cauda equina syndrome. For someone in their 80s with a history of breast cancer and new back pain we worry about metastatic spread.
There are a variety of red flags for headache and focussing your history on these is a great way not to miss anything important
As always there’s a mnemonic
(note that many of these aren’t particularly evidence based but you’re just starting in the ED so worry about the EBM a bit later)
S – Systemic symptoms:
- people often feel quite muzzy with a headache and often quite weak but if you ask specifically about things like fevers and night sweats you’ll find most don’t have them. History of cancer and immunosuppression would come as red flags in this category too
N – neurological symptoms and signs:
- funny feelings around the face are quite common in migraine but it’s certainly unusual to have unilateral limb symptoms and finding neurological signs is are even more uncommon and certainly need to be taken seriously
O – Onset:
- Two bits here.
- Age > 40 for is unusual for new headaches and should be taken seriously.
- How the headache began in terms of its speed is also important. thunderclap headache: primarily we’re interested in the speed of onset. There’s various ways to ask this and various time limits given on how sudden is sudden. I usually use “how long from the headache beginning till it was at its worst.” with some follow up questions and even some dramatic hand movements and claps to emphasise the sudden nature. (I’m a real dramatist in the history taking). Migraine and tensions headaches are often more gradual in onset. When you ask was it sudden, almost invariably patients say yes but really they’re saying this as a surrogate for severe and significant headache. They want you to take them seriously. Often when you drill down into it the headache was gradual and getting worse over several hours.
O – other associated features:
- part of this is the pattern of the headache and all the SOCRATES questions. And don’t forget the eye (glaucoma or cluster headache or a Horner’s syndrome all involve the ye) Honestly this letter isn’t as useful as the others in the mnemonic but SNOP would be a pretty crappy mnemonic…
P – Previous headache history:
- this can work both ways. The definition of migraine requires a pattern of recurrence so a recurring typical headache for them is probably a migraine and probably reassuring. However, if it was the same as all the others they probably wouldn’t be in your ED so if there’s instead a pattern of recurring worsening headaches or headaches with different features then take them seriously.
Red flag does not = CT
CT is a useful test in the right person but I suspect it reassures us and the patient and often doesn’t help a great deal with the correct diagnosis. I think the red flags should make us consider imaging and discussion with a senior can be really helpful here. But primarily the red flags should make us pause and consider our differential more seriously. For example in temporal arteritis it’s the S (systemic features) and O (age of onset) of our mnemonic that are the red flags but a CT won’t tell us anything useful here.
Normal CT does not = no pathology
The history is what dictates most of the diagnosis. While we’re doing less LPs for SAH (don’t open that pandora’s box here…) people will still have negative CT scans who have had a small SAH. CT is also useless in meningitis and even in things like dissections where you need contrast with your CT to see it. All that to say if you have just got a CT of your headache patient there’s often still some doctory work left to do.
It’s OK to ask the patient what they’re worried about
I find this really helpful for all ED patients. Some are worried about brain tumours. Some know that it’s anxiety and stress but are afraid to vocalise it (though do not dismiss headaches purely because the patient or you think they’re stressed). Once you know what they’re worried about it makes the conversation about reassurance and discharge (which is what will happen with the vast majority) so much easier.
Finally some pearls on making them feel better. Don’t wait for a definitive diagnosis to start treating your patient. Get the paracetamol and NSAID on early. Metoclopramide (sometimes in repeated doses) works for almost every kind of headache no matter what the cause (so don’t be reassured, just because the patient feels better they may still have an SAH). A litre of fluid with the metoclopramide often does wonders too (placebo or not) and if you have a clear diagnosis of migraine (and I’d get senior advice before making that diagnosis) then there’s a whole plethora of other meds that can really help.
There’s no need to panic next time you see a headache in the waiting to be seen box. The key to focus on is your history trying to tease out those red flags. And if you’re in doubt just ask.
- Clinch CR. Evaluation of acute headaches in adults. Am Fam Physician. 2001 Feb
15;63(4):685-92. Review. PubMed PMID: 11237083.