Authors: Andy Neil, Dave McCreary, Becky Maxwell, Damian Rowland, Rachel Stewart, Mark Winstanley,  / Codes: IP1, NeuP2, SLO1, SLO10, SLO12, SLO5, SLO8, XC1 / Published: 01/03/2019


The content you’re about to read or listen to is at least two years old, which means evidence and guidelines may have changed since it was originally published. This content item won’t be edited but there will be a newer version published if warranted. Check the new publications and curriculum map for updates


Andy Neill
Dave McCreary

Clinical Question:

  • Does a bilateral greater occipital nerve block (GONB) help in the treatment of migraine


  • A Randomized, Sham-Controlled Trial of Bilateral Greater Occipital Nerve Blocks With Bupivacaine for Acute Migraine Patients Refractory to Standard Emergency Department Treatment With Metoclopramide, PMID: [30144034](


  • Friedman, Headache 2018


  • Migraine is common and miserable for patients. Most never darken the door of an ED and the ones we do see are often atypical, prolonged and resistant to usual treatment with triptans. We give fluids, NSAIDs and some kind of anti-emetic such as prochlorperazine or metoclopramide. And to be honest 80-90% get better with this.
  • Ben Friedman is a headache researcher who did a lot of the clinical work on the treatments that we commonly give fro migraine in the ED. He wrote a really good “expert opinion” piece in Annals of EM a few years back and at the end of his algorithm for managing migraines he had GONB. I have been desperately wanting to try one since. In the mean time he’s gone and produced this RCT in the ED. There are other similar small trials but mostly in the clinic setting
  • The theory behind this is all to do with complicated connections between the upper cervical spinal nerves, C1&2, (the origin of the greater occipital nerve) and the massive trigeminal sensory nucleus that runs the length of the brainstem. Branches of the trigeminal and other cranial nerves are important mediators of pain either in the vessels or meninges affected in migraine. Something to do with blocking the greater occipital nerve seems to have beneficial effects on headache. Whether this is true or placebo or both we don’t know


  • RCT with sham of GONB for migraine
  • Reasonable definition and entry criteria
  • Had to need more intervention after metoclopramide to get in
  • 3mls each side of bupivacaine, sham was 0.5ml injected very superficially (which is nice as then there’s at least some superficial anaesthesia)
  • Assessors were not blinded
  • Primary outcome of pain free at 30 mins which is a relevant outcome.


  • Slow recruitment and terminated early (I suspect because metoclopramide works so damn well they couldn’t get people for the trial)
  • Meant to have 80 but only got 30
  • 0% had headache in the GONB group v 31% in the other
  • Unsurprisingly you can’t really draw firm conclusions from that


  • Damn shame this didn’t meet target but i’ll still use it if the usual treatment fails but since the I first heard of this 18 months ago I’m still waiting for someone for whom the usual treatment fails…


Dave McCreary
Andy Neill

Clinical Question:

Does IV Metoclopramide improve pain in acute migraine?

Title of Paper:

Intravenous Metoclopramide in the Treatment of Acute Migraines: A Randomised, Placebo-Controlled Trial
PMID: [30629285]
DOI: [10.1111/ane.13063]

Journal and Year:

Acta Neurologica Scandinavica January 2019

Lead Author:

Nurettin Özgür DOĞAN, M.D.


  • Migraines suck
  • There are many, many, many schools of thought on the best management of acute migraine and every department will have its own guideline. (And we all think ours is the best – a bit like shoulder relocation methods)
  • Metoclopramide is a dopamine receptor antagonist antiemetic, and may have efficacy for the treatment of the migraine itself due to its antidopaminergic features
  • Results of previous studies on its use are conflicting

Study Design:

  • Prospective, double-blind, parallel-group, placebo-controlled, randomised trial
  • Single centre, university hospital academic ED

Patients Studied:

  • Adults with migraine-type headaches as per [International Headache Society (IHS) migraine criteria](
  • As selected by a senior EM resident trained in the criteria
  • Excluded:
  • Analgesia taken in previous two hours
  • Pregnancy
  • Allergies to metoclopramide
  • Haemodynamically unstable


  • 10mg metoclopramide in 100ml normal saline IV


  • 100ml normal saline IV


  • Primary: difference in pain intensity on Numeric Rating Scale (NRS) between groups at 30 minutes
  • Secondary:
  • Adverse events
  • Need for rescue analgesic at 30th minute
  • 24-72 hours:
  • Headache intensity after discharge
  • Need for additional ED visits

Summary of Results:

  • 148 patients enrolled (this was their calculated sample size)
  • 8 left before treatment, 3 left after baseline measurements, 5 left after second measurements
  • Leaving 132 for final analysis
  • 91% had previous migraine diagnosis
  • Primary (median change in the NRS at 30 minutes): no significant difference
  • Metoclopramide: 4 (IQR 2-6)
  • Placebo: 3 (IQR 1-4)
  • Difference: -1.0 [95%CI -2.1 – 0.1)
  • When treatment success considered to be 50% reduction in symptom severity (they looked at this post-hoc):
  • Metoclopramide: 66.2% of patients
  • Placebo: 44.6% of patients
  • p=0.008 | OR 2.44 [95%CI 1.25 – 4.74] favouring metoclopramide
  • Need for rescue analgesia:
  • Metoclopramide: 21.6%
  • Placebo: 36.5%
  • Difference: 14.9% [95%CI -32.0 – 2.5]
  • Adverse affects, refractory pain and ED revisits similar between groups

Authors Conclusion:

“Overall, our findings suggest that, no difference was found between 10mg intravenous metoclopramide and placebo regarding efficacy and safety in patients with acute migraines.”

Clinical Bottom Line:

  • Despite this being a very methodically planned study, ticking a lot of EBM boxes, I still don’t quite buy it (though I’m biased, I really rate metoclopramide for migraine).
  • They technically met their sample size, but not if you consider the 11 dropouts before primary outcome measurement (that’s 7% of their sample size right there), and so I think they are probably still underpowered to truly answer the question
  • Their post-hoc analysis suggests maybe they got their primary outcome question wrong
  • Metoclopramide at a dose of 10mg may not help acute migraine…20mg, however…
  • Will this change my clinical practice? Well, no. I currently use 20mg of metoclopramide in a litre of saline and infuse over 30-60 minutes (during which time the patient has also had 900mg Aspirin, 1g of paracetamol and [12mg dexamethasone] and I find it works a treat and haven’t actually needed to go past my first line management since I settled on that little cocktail. (This is Dave Based Medicine, but I base that on the American Headache Society recommendations.

Other #FOAMed Resources / References:

  • AliEM have [this review] of prochlorperazine, metoclopramide, and diphenhydramine for acute migraine headache
  • They also have [this review] of managing migraine in the complicated patient
  • And OK [this one] isn’t FOAMed (grr) – it’s EMRAP which is behind a pay wall but if you have access Michelle Lin (off of the ALiEM review above) did a podcast segment on the American Headache Society guidelines in November 2017


Rachel Stewart
Eoghan Colgan

This podcast was recorded at EuSEM18 in Glasgow


Dave McCreary
Andy Neill

Clinical Question:

Can we estimate the probability of a UTI in febrile kids with a handy calculator?

Title of Paper:

Development and Validation of a Calculator for Estimating the Probability of Urinary Tract Infection in Young Febrile Children
PMID: [29710324]
DOI: [10.1001/jamapediatrics.2018.0217]

Journal and Year:

JAMA Pediatrics 2018

Lead Author:

Nader Shaikh MD MPH


  • Reportedly around 7% febrile kids <2 years in the ED have a UTI
  • Getting using to test for UTI in this age group is not the easiest, and can include the need for an in-out catheter
  • Wouldn’t it be nice to know which kids really need us to go to these lengths?
  • Even when we do catch wee, we still need to decide based on the results (say just a trace of leucs) who need empirical treatment vs wait for the culture
  • We should be trying to reduce unnecessary antibiotic use, where safe and possible
  • The authors have developed UTICalc – a calculator that first estimates pretest probability based on clinical variables, and updates post-test probability based on lab tests, if performed
  • This study is the derivation and validation of this calculator


  • Retrospective chart review of separately created training (derivation) database (1686 patients) and validation database (384 patients).
  • Febrile children <2 years old presenting to a Paediatric ED over study period (2007-2013)
  • All patients with confirmed UTI (n=570)
  • Random sample of patients without UTI (n=1312)


  • Developed 5 multivariable logistic regression models to estimate risk of UTI and produced ROC curves for each
  • One clinical model with 5 clinical risk factors: age <12 months, temp ≥39oC, nonbank race, female or uncircumcised male, no other source identified
  • Four laboratory models: dipstick, dipstick + gram stain, dipstick + urine WCC (hemocytometer), and an enhanced urinalysis model of variables from clinical, hemocytometer and gram stain.


  • With AUC they aimed for 95% sensitivity
    – 2% cutoff for clinical model – the point at or above which patient has high pretest probability for UTI and those requires urine testing
    – 5% cutoff for lab models – the point at or above which patient has relatively high post-test probability of UTI and should be treated with antibiotics


  • Chart review of patients <2 years presenting to the ED with temp ≥38oC to test performance of the UTICalc

The calculator:

Summary of Results (Validation):

  • 384 patients
  • 60% ages 2-11 months
  • 76% female


  • If used on a population of 1000 patients, compared to the [American Academy of Paediatrics algorithm], UTICalc would have:
  • Reduced the need for urine sampling by 8.1% [95%CI 4.2-12%]
  • Reduced missed UTI from 3 cases to 0
  • Reduced number of delayed treatment by 10.6% [0.9-20.4%] without substantially increased antibiotic use, based on the dipstick component (as in before the lab result has returned)

Authors Conclusion:

Accurate diagnosis of UTI is important to reduce the delay in diagnosis and to avoid unnecessary treatment with antimicrobial drugs. The approach advocated here tailors testing and treatment to the risk factors present in the child being assessed, thus offering the potential to improve outcomes with UTI.

Clinical Bottom Line:

I like clever calculators like this, and this is an impressive piece of work in my humble opinion.  I think the can really add some objective numbers to our clinical assessment if we use them correctly.

Like always with a newly internally validated tool such as this, it needs tested in external populations before it can truly be relied upon.

Other #FOAMed Resources / References:

  • Blast from the past – don’t forget the [Quickwee method] when it comes to catching that elusive pee.  You can hear/read our NewInEm review of it [here]
  • [Journalfeed] reviewed this paper [here]
  • The calculator itself can be found [here]

Gillick competence relates to whether a child can consent or not. This determines that children under 16 can consent if they have sufficient understanding and intelligence to fully understand what is involved in a proposed treatment, including its purpose, nature, likely effects and risks, chances of success and availability of other options. This means children can be deemed “gillick competent” and consent to treatment. If they are not gillick competent, the consent of a person with parental responsibility is needed.

Fraser guidelines are specifically relating to contraceptive advice, sexuallly transmitted infections and terminating of pregnancy. It is less likely we will need these in the ED. These guidelines look at the child’s maturity and intelligence, whether they can be persuaded to tell their parents, if they will continue sexual intercourse with or without contraception, physical or mental health is likely to suffer unless they receive advice or treatment, and the advice or treatment is in the young person’s best interests.

There is no lower age limited for these guidelines, but it is rarely considered approrpiate for under 13 year olds to consent to any treatment.

Cauda Equina Guideline


Becky Maxwell and Chris Connolly