Author: Dr. Calloway Pichette
Background: Headache, a common complaint in the ED. Many times a day we see patients with migraine headaches in the ED. Give them a migraine cocktail (sadly, no martinis involved…) and then viola! They feel better and they’re on their way. But do we have to give all migraine patients IV fluids? Does it make that much of a difference in their pain scale and outcome? That’s what this study aimed to assess. It’s important to note that this was a pilot study, with the aim of seeing what the outcomes were to see if a large scale study asking this question was feasible/would provide useful information.
Design: This was a blinded, single center study that took place in an urban teaching hospital in Camden, NJ. Patients were screened for inclusion 8:30 a.m.-10:30 p.m. 7 days per week. Patients were eligible for inclusion if they were 18 years of age, had a headache meeting the third international classification for headache disorders definition of a migraine headache, and had a history of 5 prior similar headaches. Exclusion criteria included those who were pregnant, non-english speaking, received more than 500 mL of fluid prior to enrollment, or if the ED doc determined that fluids were definitively indicated or contraindicated.
Patients were randomized to either receive a 1 L bolus of normal saline or 10 mL/hr of normal saline. All patients received 10 mg IV prochlorperazine and 25 mg of diphenhydramine. Everyone involved in the study except the bedside nurses were blinded to which treatment arm the patients were in. (of note, they said they put a dark bag over the fluids so they patients didn’t know what they were getting.) Patients were assessed on pain, nausea, and functional disability at 0, 60, and 120 minutes, will a phone call 48 hours after the study for additional follow up.
Results: The primary outcome was a difference in pain scale (1.3 is clinically significant). Secondary outcomes were the difference in 0 and 120 minute pain scales, pain free status at 120 minutes, functional disability and nausea at 60 and 120 min, 48 hour pain scale, global assessment at 48 hr, need for rescue medications, and the feasibility of their protocol for a large scale randomized controlled trial.
In total, they had 50 patients, with 25 in each arm. One person in the control arm withdrew from the study, but were analyzed according to intention to treat analysis. The randomization worked out pretty well and the patient characteristics were quite similar between the two groups (woo!). The paper goes into lots of statistics and breaks down all numbers for the outcomes they were measuring, but the main important point is that the change is pain is not statistically significant between the two study groups. (Difference between fluid bolus and control at 60 minutes with the 95% CI was -0.4 (-1.9 to 1.1) and at 120 minutes was 0.5 (-1.3-2.2).)
More statistical wizardry and simulations revealed that 65 persons per group would be needed to to get a P values of 0.05 for 80% power, with 85 per group with 90% power if they were to do a larger scale trial. Based on simulations, the authors concluded that a large scale trial was unlikely to show a clinically important treatment effect caused by the intervention.
Discussion: This study demonstrated that IV fluids don’t result in a statistically significant effect when it comes to decreasing pain with migraine headache, however, the clinical effects may be significant. Many people with migraines are dehydrated, and dehydration can make the perception of pain worse and contribute to the migraine headache itself. Not to mention, patients seem to think there is something magical about IV fluids. I can’t tell you them number of times I’ve heard “you’re not going to give me fluids?” with a disappointed tone when, for some reason, I wasn’t planning on it. I’m not one to diss the placebo effect because hey, it’s one hell of a drug.
If there’s no reason for me not to give IV fluids to a patient with migraine headache, I’ll continue to do so because it might help them feel better and may give a boost of that placebo effect to my side. That being said, if we can’t get IV access on a patient or there’s a reason they can’t get fluids, I won’t stress about not being able to give them IV fluids because we know it doesn’t make that much of a difference. If there’s no major contraindication to them drinking a large amount of water, I’ll just set a large thing of water in the room with the patient and tell them to drink up.
For a pilot study I thought this was well designed and they stayed consistent at what they were looking at, with lots of pre-determined secondary outcomes. They blinded everyone they could and used intention to treat analysis. They said they asked patients if they knew how much fluid they got (I would think I’d know if I was getting a liter of fluids even if it was covered in a black plastic bag) but they said only 1 person guessed correctly which treatment group they were in. Either people are really bad about guessing, or they study personnel hid the bags in a way that actually made it hard to tell (like all bags of the same size all on an IV pump or something.) Obviously, being a small pilot study in one center introduces lots of limitations. The study was also limited on length of stay data because of their predetermined 120 minute assessment cut off. I also don’t expect to see any large scale studies looking at this question to be published soon because the results of this pilot study indicate they likely wouldn’t be that different.
Citation: Christopher W. Jones, Lauren B. Remboski, Brian Freeze, Valerie A. Braz, John P. Gaughan, Samuel A. McLean,Intravenous Fluid for the Treatment of Emergency Department Patients With Migraine Headache: A Randomized Controlled Trial,Annals of Emergency Medicine,Volume 73, Issue 2, 2019, Pages 150-156, ISSN 0196-0644, https://doi.org/10.1016/j.annemergmed.2018.09.004. (http://www.sciencedirect.com/science/article/pii/S0196064418312629)