PODCAST: Hit Me with Your Best Block – 2025 AHS ED Migraine Guidelines

Podcast: Play in new window | Download

Date: January 5, 2026

Reference: Robblee et al. 2025 guideline update to acute treatment of migraine for adults in the emergency department: The American Headache Society evidence assessment of parenteral pharmacotherapies. Headache 2025 Dec

Happy New Year, SGEMers! What better way to start 2026 than with an SGEM Xtra about migraine headaches? We were originally scheduled to record this episode in December, but circumstances changed.

This is another SGEM Xtra and not the typical structured critical appraisal with a checklist. It will be a conversation about what we should be doing and should stop doing when treating migraine patients in the ED based on the new American Headache Society (AHS) guidelines. However, you will find a standard SGEM nerdy critical appraisal at the end of this blog post.

Migraine is one of the most common causes of headache visits to the ED, representing ~¼ of the 3.5 million annual headache-related visits in the US. Despite prior guidelines, ED practice is still all over the map, and patients sometimes leave without much relief. The AHS has just released the 2025 guideline update on parenteral pharmacotherapies and nerve blocks for adult ED migraine. To help us understand these new guidelines, we are joined by two neurologists who literally wrote the guidelines.

Dr. Jennifer Robblee

Dr. Jennifer Robblee (lead guideline author) is a Board‑certified neurologist and headache specialist at Barrow Neurological Institute in Phoenix. Her practice focuses on refractory migraine and status migrainosus. She trained at the University of Toronto (MD, neurology residency, MSc) and completed a headache fellowship at the Mayo Clinic Scottsdale.

Jennifer is the third eurologist to be on the SGEM. We’ve had Dr. Jeff Saver and Dr. Ravi Garg discuss thrombolytics and stroke. This will be an example that not all of neurology and emergency medicine intersect over stroke care.

Dr. Serena Orr

Dr. Serena Orr (senior guideline author) is a pediatric neurologist, headache subspecialist, and director of the pediatric headache program at Alberta Children’s Hospital in Calgary. Serena has a strong interest in acute treatment of migraine, tech‑based treatment solutions, and psychosocial factors affecting migraine in kids and teens.

The AHS guideline committee uses a 5-year update cycle for guidelines. Since 2016, 26 new RCTs and 20 injectable treatments, including nerve blocks (GONB, SONB, SPG) and eptinezumab.  Unfortunately, ED migraine outcomes are still not great. Only ~37% of ED patients achieve headache freedom at discharge.

These new guidelines were trying to answer two questions.

1. Which injectable meds are effective in adults with migraine in the ED?
2. Are nerve blocks effective in adults with migraine in the ED?

Top 5 things ED should know about the 2025 AHS Migraine Guidelines

Listen to the SGEM podcast to hear Jennier and Serena discuss the top five things emergency physicians should know about the 2025 migraine guidelines.

1. Prochlorperazine IV & Greater Occipital Nerve Blocks (GONB) Are Now Level A “Must Offer”

• IV prochlorperazine and greater occipital nerve blocks (GONB) are Level A – must offer or adults presenting to the ED with a migraine attack requiring parenteral therapy (if no contraindications).

• Questions:
  • This is a big upgrade from 2016. Why did prochlorperazine and GONB earn Level A status in 2025?
  • Practically, what does that look like in an ED order set? Are you imagining that everyone gets prochlorperazine?
  • For the EM docs who have not been performing occipital nerve blocks, how steep is the learning curve?

2. Hydromorphone Is Level A “Must NOT Offer”

• Hydromorphone IV: Level A – Must NOT offer for migraine in the ED.

• Questions:
  • Let’s talk about opioids. Hydromorphone is now ‘must NOT offer’, what tipped the scale to Level A harm/no benefit?
  • “Must NOT offer” seems like a strong statement (thou shalt not), is there not a potential clinical situation where an opioid still should be offered?
  • How do we balance real‑world pressures, patient expectations, throughput, Press Ganey scores with an anti‑opioid, evidence‑based stance? It’s going to impact ED docs and not neurologists.

3. The Level B Recommendations: 

• Level B – “Should offer” for headache requiring parenteral therapy (Dexketoprofen IV, ketorolac IV, metoclopramide IV, subcutaneous sumatriptan, and supraorbital nerve blocks [SONB]).
• Dexamethasone IV remains Level B “should offer” for recurrence prevention from the 2016 guidance.

• Questions:
  • If Level A is your starting lineup, who’s on the bench as your Level B ‘should offer’ options, and when do you pull them in?
  • Is there a preferred sequence – dopamine antagonist first, then NSAID, then triptan, or is it more patient‑specific?
  • How should ED clinicians think about dexamethasone? Is it still a routine add‑on, or more selective?

4. Nerve Blocks Are Mainstream

• GONB: Level A – Must offer.
• SONB: Level B – May/should offer when GONB is insufficient or not possible.

• Questions:
  • For a busy ED, how realistic is it to integrate occipital and supraorbital nerve blocks into standard migraine care?
  • What’s the pragmatic advice on training?  
  • Can EM doctors become competent with blocks via bedside teaching and FOAMed resources>

5. Big Evidence Gaps

No meta‑analyses were possible because of significant heterogeneity in methods and outcomes. Additional ED-specific outcomes, such as pain relief at 1 hour. Asking about patient-oriented outcomes (POO) such as “Would you want this treatment again on your next ED visit?” Need ED‑specific data on eptinezumab (currently Level U for general ED use despite strong outpatient data).

• Questions:
  • If you had unlimited funding for one ED migraine randomized control trial, what would you test, and what outcome would you choose?
  • You recommended a 1‑hour endpoint for ED trials. How does that change how we design and interpret future studies?
  • I love the idea of the patient-centred outcome: ‘Would you want this again?’ How do we make sure future trials include that kind of measure?

Five Limitations of the AHS Migraine Guideline

The goal here is not to dunk on the guideline; there are limitations to any study.  This is just a nerdy conversation about how the next cycle could be improved. Listen to the SGEM Xtra podcast to hear Jennifer and Serena respond.

Limitation 1: Risk of Bias Tool & Study Quality Nuances

• Question: “You explicitly say that some ‘class I’ RCTs had small sample sizes or weird time points that made you less confident. From an EBM standpoint, how did you reconcile the AAN RoB categories with what we’d call imprecision and indirectness in GRADE?”

Limitation 2: External Validity – Not All RCTs Were ED RCTs

• Question: “Many of the trials you had to work with weren’t actually done in ED settings – eptinezumab and SPG blocks being two examples. How worried should we be about extrapolating outpatient data into the ED, where patients are often later in the attack, more distressed, and maybe have different comorbidities?”

Limitation 3: Active Comparators of Unclear Significance

• Question: “You call out trials that compare against ‘iffy’ active controls (valproate, dexamethasone, etc). In EBM terms, this muddies the signal. If you beat a weak comparator, is your drug actually good? How did you handle that when grading evidence and crafting recommendations?”

Limitation 4: No Meta‑Analyses; Reliance on Narrative Synthesis

• Question: “From a methodological point of view, the fact that you couldn’t meta‑analyze anything limits precision and makes it hard to quantify effect sizes. How should EM clinicians interpret Level A or B recommendations that rest on narrative synthesis instead of pooled estimates?”

Limitation 5: Broader Biases – Publication, Selection, and the ED Reality

• Question: “Zooming way out, every guideline sits on top of the published RCT iceberg. How much do you worry about publication bias, selection bias, and the fact that ED patients we see at 3 am rarely look like the trial population?”

Remember to be skeptical of anything you learn, even if you heard it on the Skeptics’ Guide to Emergency Medicine. 

Critical Appraisal

Reference: Robblee et al. 2025 guideline update to acute treatment of migraine for adults in the emergency department: The American Headache Society evidence assessment of parenteral pharmacotherapies. Headache 2025 Dec

Background: Migraine is one of the most common reasons people roll into the ED with a headache, and it’s not just “a bad headache.” It’s a chronic neurologic disorder that affects over a billion people globally and is consistently among the top causes of years lived with disability, especially in young and middle-aged adults [1.2].  In the ED specifically, migraine accounts for about one‑quarter of the ~3.5 million headache-related visits per year in the US. That’s a lot of stretchers tied up with photophobic patients in dark rooms.

Clinically, migraine is defined by the International Classification of Headache Disorders (ICHD‑3). Typical attacks last 4 to 72 hours and are moderate to severe, often unilateral, pulsating, and worsened by routine physical activity. They’re commonly accompanied by nausea and/or vomiting and photophobia/phonophobia [3]. Migraine without aura is the most common type; migraine with aura adds transient focal neurologic symptoms (usually visual) that precede or accompany the headache. Diagnosis in the ED is clinical: apply ICHD‑3 criteria, look for a typical migraine phenotype, and screen for red flags (fever, meningeal signs, focal deficits, thunderclap onset, immunocompromise, anticoagulation, etc.) to rule out secondary causes.

Outside the ED, acute migraine treatment usually starts with oral NSAIDs or acetaminophen, triptans, and newer agents like gepants or ditans, often combined with antiemetics. Preventive therapy (beta‑blockers, topiramate, CGRP monoclonal antibodies) targets attack frequency and disability, not the single ED visit [4,5].  In the ED, however, patients usually present with moderate–severe attacks that have failed home therapy and can’t tolerate oral meds, so parenteral therapies (IV/IM/SC drugs and nerve blocks) dominate practice. Historically, ED care has been all over the map, with substantial opioid use and only about 37% of patients leaving headache‑free in one large study. This is the gap the new American Headache Society (AHS) guideline is trying to address.

Clinical Question: In adults presenting to the ED with an acute migraine attack requiring parenteral therapy, which injectable medications and nerve blocks improve pain and related outcomes, and which should be avoided?

• Population: Adults (>18 years) with an ICHD-diagnosed migraine who presented to the ED with a migraine attack. The update includes 26 new randomized controlled trials (RCTs) with 3019 additional participants, added to the 2016 evidence base. Most trials enrolled adults with moderate–to–severe migraine, often with previous migraine history; some nerve‑block and biologic trials included more selected or outpatient populations, which becomes a limitation later.
  • Exclusions: Children, non‑ED settings, non‑parenteral interventions or comparators, nonmigraine headaches, non‑RCTs, and non‑English trials.
• Interventions:
  • Medications: Dopamine antagonists (prochlorperazine IV, chlorpromazine IV, metoclopramide IV, haloperidol IV/IM, and droperidol IM), NSAIDs and related (ketorolac IV, dexketoprofen IV, diclofenac IM, IV aspirin, dipyrone IV, and ibuprofen IV), Antiepileptic (valproate IV), Corticosteroid (dexamethasone IV), Antihistamine (diphenhydramine IV – mainly for akathisia prevention and as an active comparator), Opioids (hydromorphone IV, morphine IV, meperidine IV, nalbuphine IV, tramadol IV), Other agents (magnesium IV, caffeine IV, granisetron IV, ketamine IV, lidocaine IV, paracetamol/ acetaminophen IV, normal saline IV, propofol IV, dihydroergotamine IV/SC, octreotide IV/SC, ergotamine SC, lysine clonixinate IV, trimethobenzamide IM and biologic (eptinezumab IV).
  • Nerve Blocks: Greater occipital nerve block (GONB), supraorbital nerve block (SONB) and sphenopalatine ganglion (SPG) blocks. These are delivered with local anesthetics like lidocaine or bupivacaine in various doses and volumes.
• Comparison: Placebo parenteral injections (saline or sham blocks) or active parenteral comparators (another ED migraine drug or a ess well‑supported agents like IV valproate or dexamethasone used as controls). Non‑parenteral comparators (oral agents) were explicitly excluded.
• Outcomes: Each trial had to report pain outcomes within 6 hours of administration.
  • Primary Outcomes: Change in headache intensity on a numerical rating scale (NRS) or VAS at an early time point (often 1–2 hours). Proportion of patients achieving “headache freedom”or at least 50% reduction in pain by a specified time (often discharge/2 hours). Need for rescue medication in the ED
  • Secondary Outcomes: Headache recurrence after discharge (24 to 48 hours). Functional outcomes (ability to resume normal activity). Patient‑reported satisfaction or willingness to receive the same treatment again. Adverse events (akathisia, dystonia, hypotension, sedation, local anesthetic toxicity)

Authors’ Conclusions: “Prochlorperazine IV and GONB must be offered to eligible adults presenting to the ED with a migraine attack for treatment of headache requiring parenteral therapy (level A – must offer) in those without contraindications, while hydromorphone IV must not be offered (level A – must not offer). Treatments that should be offered when appropriate (level B – should offer) include dexketoprofen IV, ketorolac IV, metoclopramide IV, sumatriptan SC, and SONB. Chlorpromazine IV, dexamethasone IV, and valproate IV may be offered (level C – may offer). Paracetamol IV may not be offered (level C – should not offer). Eptinezumab should be offered (level B) only for patients matching the clinical trial population but is rated level U – no recommendation for an ED-­ specific population. Additional evidence is needed for caffeine, granisetron, ibuprofen, ketamine, lidocaine, normal saline, propofol, and SPG blocks, all currently rated level U – no recommendation.”

Quality Checklist for Guidelines (Yes/No/Unsure)

1. Study population included or focused on those in the emergency department? Yes
2. Explicit and sensible process used to identify, select and combine evidence? Yes
3. Quality of the evidence explicitly assessed using a validated instrument? Yes
4. Explicit and sensible process used to weigh the relative value of different outcomes? Unsure
5. Guideline thoughtfully balances desirable and undesirable effects? Yes
6. Guideline accounts for important recent developments? Yes
7. Guideline has been peer‑reviewed and tested? Yes/No
8. Practical, actionable and clinically important recommendations are made?
   Yes
9. Guideline authors’ conflicts of interest fully reported, transparent and unlikely to sway recommendations? Unsure

The guideline uses the AAN/AHS scheme for communicating the strength of recommendations:

• Level A – Must / Must NOT offer
• Level B – Should / Should NOT offer
• Level C – May / May NOT offer
• Level U – No recommendation

Key Recommendations:

Level A – Must Offer & Must NOT Offer

• Must offer (Level A):
  • Prochlorperazine IV for adults with migraine attacks in the ED requiring parenteral therapy, without contraindications.
  • Greater occipital nerve block (GONB) with local anesthetic as an effective acute treatment in ED patients with migraine.
• Must NOT offer (Level A):
  • Hydromorphone IV is judged likely ineffective and potentially harmful, and therefore recommended against for migraine-related pain relief.

Level B – Should offer / Should NOT offer

• Should offer (Level B – positive): For adults with migraine in the ED requiring parenteral therapy, when clinically appropriate (Dexketoprofen IV, Ketorolac IV, Metoclopramide IV, Sumatriptan SC and Supraorbital nerve block)
• Should offer (Level B – special case): Eptinezumab IV (100 mg) should be offered only to patients who closely match the outpatient clinical trial population; for general ED use it is rated Level U (no recommendation) because ED‑specific evidence is lacking.
• Should NOT offer (Level B negative): No new designations at Level B negative beyond existing opioids, being strongly discouraged overall. The clearest negative recommendation is hydromorphone at Level A, which must not be offered.

Five Limitations that Threaten the Validity of the Guidelines:

1. Heterogeneity: Although the authors planned meta-analyses when ≥2 class I/II studies with comparable methods were available, they appropriately did not perform any meta-analyses because no intervention met the prespecified criteria for pooling. This means the evidence synthesis is entirely narrative, relying on qualitative judgment rather than quantitative pooling. Guidelines based on narrative synthesis are inherently more fragile than those built on robust, transparent meta-analyses where possible.
2. Use of Active Comparators with Uncertain Efficacy: Several RCTs used active controls such as IV valproate or dexamethasone. These are agents whose own efficacy is not strongly established in this setting and could be considered a “strawman”. The guideline explicitly flags this as a limitation, noting that “active controls of unclear significance” complicate interpretation, and that it can be unclear whether similar outcomes indicate both treatments are good or neither is effective. Trials using inadequately validated comparators risk biocreep and make effect sizes difficult to interpret. These comparator issues propagate up into guideline grading and can threaten the internal validity of some recommendations.
3. Limited ED‑Specific Data: The guideline includes evidence from non-ED populations for some interventions, notably eptinezumab IV and commercial SPG block kits, and appropriately downgrades or withholds ED-specific recommendations. From a validity standpoint, importing outpatient/chronic data into an ED acute‑care guideline raises concerns about indirectness (PICO mismatch), a key downgrade domain in systems like GRADE. This weakens the strength and applicability of those specific recommendations to ED patients.
4. Conflict of Interest & Potential Editorial Influence: The guideline authors fully disclose substantial relationships with industry (honoraria, advisory roles, research funding, and equity positions). While transparency is great and consistent with IOM standards, both the IOM and later methodological reviews stress that trustworthy guidelines should minimize and manage conflicts, particularly among chairs and voting members, given evidence that financial ties can shift recommendations toward more favourable or aggressive treatment. The paper notes reliance on AAN/AHS procedures and consensus when evidence was limited, which is precisely where COI‑related bias can creep in (choice of thresholds, downgrading vs not, enthusiasm for newer agents). We don’t have evidence that bias occurred, the degree of industry involvement should make us more skeptical of the recommendations.
5. Patient Values & Preferences: Multiple widely used guideline-development standards explicitly recommend involving patients/public (IOM, WHO, AGREE II, NICE and GIN). Knowing what patients want is one of the three pillars of evidence-based medicine (EBM). These guidelines did not involve any patient input during their development. It would be great if, in the next guideline, the AHS could find a way to engage with patients and ensure their preferences and values are known.

SGEM Bottom Line: For adults with acute migraine in the ED, intravenous prochlorperazine and greater occipital nerve blocks are now “must‑offer” treatments, opioids (especially hydromorphone) and IV acetaminophen should be avoided, and many other parenteral agents remain either “should offer,” “may offer,” or “no recommendation” based on a heterogeneous but systematically appraised evidence base.

References:

1. Dong L, Dong W, Jin Y, Jiang Y, Li Z, Yu D. The Global Burden of Migraine: A 30-Year Trend Review and Future Projections by Age, Sex, Country, and Region. Pain Ther. 2025 Feb;14(1):297-315. doi: 10.1007/s40122-024-00690-7. Epub 2024 Dec 11. PMID: 39661241; PMCID: PMC11751287.
2. GBD 2023 Headache Collaborators. Global, regional, and national burden of headache disorders, 1990-2023: a systematic analysis for the Global Burden of Disease Study 2023. Lancet Neurol. 2025 Dec;24(12):1005-1015. doi: 10.1016/S1474-4422(25)00402-8. PMID: 41240916; PMCID: PMC12612381.
3. HIS Classification of ICHD-3: https://ichd-3.org/1-migraine/1-1-migraine-without-aura/ Accessed December 5, 2025
4. Pescador Ruschel MA, De Jesus O. Migraine Headache. [Updated 2024 Jul 5]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2025 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK560787/
5. Medscape. Chawla et al. Migraine Headache. https://emedicine.medscape.com/article/1142556-overview?form=fpf  Accessed