Written by: Vidya Eswaran, MD (NUEM PGY-2) Edited by: Danielle Miller, MD, (NUEM PGY-3) Expert commentary by: Seth Trueger, MD, MPH
Thanks for this great overview over ED headache management. Our approach to headaches has matured in recent years, largely because of a bunch of great studies by Ben Friedman’s group at Montefiore (COI: his brother and I were residency classmates); see Headache guidelines (he’s the first et al in the Orr paper cited above, REF); his Annals Expert Clinical Management paper [REF]; and his FOAM post at ALiEM.
My general approach to headaches:
First: Is there a dangerous cause?
Is the headache similar to their prior headaches in character, location, magnitude, timing, associated symptoms? Are there concerning features (exertional, vomiting, personal or family history of aneurysms)? Was it maximal at onset (or sudden/severe)? I find it helpful to ask: “what were you doing when it started?”; “how bad was it when it started?”; “when was it the worst?”; and only after listening for a while, I backdoor into whether it is typical for them (e.g. “it’s usually on that side and you’re nauseated when you get a headache like this…?”). I’ve gotten myself in trouble by asking up front if it’s the same as usual or “worst headache of your life” – even if they don’t mean to, patients sometimes seem like they are trying to validate why they came to the ED (to us or to themselves). Of course, none of these questions are black and white and there’s a lot of room for clinical judgment; one minor deviation from typical headache does not mandate imaging. Patients come to see us for our expertise and often find it reassuring that we’ve listened and examined them and aren’t concerned.
Second: Symptom management
Turn off the lights
Even migraines without frank photophobia often feel better in a darker room. I usually turn off the lights as soon as I walk in. No reason to wait.
Any of the dopaminergic antiemetics are effective; I generally use whichever is typically used in my ED (and doesn’t have to come from pharmacy). IM works fine if the patient doesn’t have an IV – most headache patients don’t need labs or IV fluids so no reason to start one routinely and a lot of even severe headaches are suitable for fast track. But they just don’t seem to work PO.
I make sure the patient is up to their appropriate daily dosing of acetaminophen or ibuprofen; with appropriate consideration of contraindications, they might make a difference so why not? PO ibuprofen is likely as effective as ketorolac, and getting a shot doesn’t seem to have a placebo effect (if nothing else, this study is worth reading for the amazing design [REF]; summarized in an accompanying editorial on placebos [REF]).
Steroids don’t fix the headache today but they decrease recurrence in some patients. I don’t give them to everyone, but for patients who get headaches in groups, have been having headaches for a while, or are just miserable enough, I give 10mg of dexamethasone.
Diphenhydramine doesn’t work for headaches [208 patient RCT, REF].
Diphenhydramine doesn’t prevent metoclopramide-associated akathisia [REF; REF] (which in my experience, is much less common than the literature describes). Midazolam has some effect for prophylaxis [REF], but the rate of akathisia is low enough that I don’t think it’s worth the risks (or extending LOS due to zonking out the patient). [n.b. the same for avoiding prophylaxis for ketamine sedations; REF] If the patient gets akathisia [or an emergence reaction], then I give midaz.
I’ve heard people suggest diphenhydramine works by knocking out the patient and as any migraine sufferer knows, the best treatment is probably sleep, but the evidence suggests that diphenhydramine just doesn’t add much.
That being said, I pick my battles and I don’t knock it our of nurses hands or reprimand the residents every time or even fight with patients if they really think it helps. But never push IV Benadryl – it gets you high [REF].
Unless the patient has been vomiting a lot or has another reason to be volume down, there is little reason to give IV fluids [REF], and in my experience, it locks the patient’s LOS into at least however long the bag takes to drip in, and of course patients who need fluids the least are most likely to have the most positional IVs….
I’ll be honest, I don’t give triptans, except in the rare cases where patients know they work for them. Usually the patient’s already taken their home dose, and it seems like most patients have missed the window for effectiveness by the time they’re seeing me. I admit that I’m behind the science and this probably has more to do with practice patterns during my training. I’m open to being convinced.
If the patient falls asleep and my ED has the bandwidth, I rarely wake them up. Seems worth giving up the room for a few hours for what’s essentially curative therapy for a miserable condition.
Occipital Nerve Block
If the headache is occipital and radiates forward at all, it could be occipital neuralgia, which often responds quite well to occipital nerve blocks. The diagnosis is not very scientific, but the potential treatment is very simple and, if nothing else, just takes a few minutes of my time and a little pain for the patient, which might provide them hours of relief. I recommen a bupivicaine occipital nerve block to any patient who I think could have occipital neuralgia, review the anatomy, make sure I'm not in the artery, and put in a nice dose of bupivicaine.
Opioids don’t work for headache. I never* give opioids for headache. Don’t give opioids for headache.
That being said, the evidence base is surprisingly weak (see REF). My personal experience resonates here; I’ve had a handful of classic migraines, and some were when I was studying abroad in Australia where codeine is OTC. It made me sleepy-ish but didn’t help the headache at all, for whatever that’s worth.
*Rarely, if I see a patient with a pain contract that includes opioids for headaches from a reputable source, I don’t die on that hill.
2nd round: magnesium + (metoclopramide or haloperidol)
If the patient need a second round (or if their headache was terrible to begin with) I throw some IV mag at them. The evidence is weak at best, but it might help and is pretty safe, so why not.
I usually re-dose metoclopramide at this point, but if their current or prior headaches are generally refractory, I often switch to haloperidol (2.5mg IV or IM).
Key points in communicating with the patient and evaluating them for discharge:
I don’t have a silver bullet to fix their headache. My goal is to make sure we’re not worried that there is something dangerous going on (“good news! we’re not”) – it’s safe to go home and we have a good, safe plan for follow up (we’re not just kicking them out). While I can’t make the headache go away completely, “my other goal is to get you to the point where you can be miserable here or miserable at home” and we can safely discharge with return instructions and follow-up; I consider Neuro or headache specialist referral if it seems appropriate.
I really think that articulating a lot of these steps is helpful. Much of what we do implicitly is not clear to the patient – our across-the-room gestalt, our assessment and thought processes. Patients want to be listened to, they want to know what to expect, they want to know what to do, and they want a doctor who cares.
Seth Trueger, MD, MPH
Assistant Professor, Northwestern Emergency Medicine
How to cite this post
[Peer-Reviewed, Web Publication] Eswaran V, Miller D (2018, April 30 ). The Migraine Cocktail: Emergency Department Management of Headaches. [NUEM Blog. Expert Commentary by Trueger, S ]. Retrieved from http://www.nuemblog.com/blog/headache