Is greater occipital nerve block with local anaesthetic effective for the treatment of acute migraine in the emergency department
Date First Published:
July 17, 2026
Last Updated:
July 17, 2026
Report by:
Jackie Tsang, Fourth Year Emergency Medicine Resident (University of Toronto)
Search checked by:
Sara Vaughan, FRCPC, Fifth Year Emergency Medicine Resident
Three-Part Question:
In [adults presenting to the emergency department with an acute migraine], does a [greater occipital nerve block with local anesthetic] , [improve acute pain outcomes]?
Clinical Scenario:
You are working in a busy emergency department when a 34-year-old woman presents with a severe acute migraine. She has a well-established diagnosis of episodic migraine and has already taken a triptan and NSAID without relief. She is photophobic and vomiting, and is reluctant to receive intravenous medications because of a previous adverse reaction to metoclopramide. You wonder whether a greater occipital nerve block could improve her acute pain relief.
Search Strategy:
A database search was completed on June 16th 2026, using MEDLINE (1974 – 2026) via the Ovid interface with the following search terms
# Block 1: PATIENT / PROBLEM
exp Migraine Disorders/ OR "migraine".mp. OR "acute migraine".mp. OR "migraine attack".mp. OR "migraine headache".mp.
# Block 2: INTERVENTION
exp Nerve Block/ OR "greater occipital nerve block".mp. OR "occipital nerve block".mp. OR "GONB".mp. OR "peripheral nerve block".mp. OR "occipital nerve blockade".mp.
# Block 3: OUTCOMES
"pain relief".mp. OR "pain score".mp. OR "visual analogue scale".mp. OR "numeric rating scale".mp. OR "VAS".mp. OR "NRS".mp. OR "headache relief".mp. OR "headache freedom".mp.
# Combine
1 AND 2 AND 3
# Limits
limit to (humans and English language and yr="2009-current")
Following this, a supplementary search was conducted using the Google Scholar ‘cited by’ function to identify additional studies that had referenced our eligible papers. Additionally, the reference list of each paper was screened to identify any studies missed by the Ovid search.
# Block 1: PATIENT / PROBLEM
exp Migraine Disorders/ OR "migraine".mp. OR "acute migraine".mp. OR "migraine attack".mp. OR "migraine headache".mp.
# Block 2: INTERVENTION
exp Nerve Block/ OR "greater occipital nerve block".mp. OR "occipital nerve block".mp. OR "GONB".mp. OR "peripheral nerve block".mp. OR "occipital nerve blockade".mp.
# Block 3: OUTCOMES
"pain relief".mp. OR "pain score".mp. OR "visual analogue scale".mp. OR "numeric rating scale".mp. OR "VAS".mp. OR "NRS".mp. OR "headache relief".mp. OR "headache freedom".mp.
# Combine
1 AND 2 AND 3
# Limits
limit to (humans and English language and yr="2009-current")
Following this, a supplementary search was conducted using the Google Scholar ‘cited by’ function to identify additional studies that had referenced our eligible papers. Additionally, the reference list of each paper was screened to identify any studies missed by the Ovid search.
Search Details:
Studies were included if they met the following criteria: Randomised controlled trials, prospective controlled studies, and retrospective cohort studies; adults (≥18 years); ICHD-defined acute migraine; GONB with local anaesthetic (with or without comparator); primary outcome of acute pain score change.
Studies were included if they met the following exclusion criteria: Studies of GONB exclusively for migraine prophylaxis; paediatric populations; non-English language; studies using GONB with corticosteroid as the primary agent without a local-anesthetic-only arm.
Studies were included if they met the following exclusion criteria: Studies of GONB exclusively for migraine prophylaxis; paediatric populations; non-English language; studies using GONB with corticosteroid as the primary agent without a local-anesthetic-only arm.
Outcome:
Of the 100 papers identified, 93 were excluded after title, abstract, and full-text review. Papers were most commonly excluded because they addressed chronic or preventive migraine management rather than acute ED treatment (n=32), or because they evaluated a different intervention not meeting the GONB-with-local-anaesthetic criterion (e.g., sphenopalatine ganglion block, stellate ganglion block, occipital nerve stimulation, botulinum toxin, or IV/intranasal lidocaine (n=27). Additional exclusions included studies involving non-migraine headaches, incorrect indications, basic science studies, narrative reviews without GONB or ED-specific data, inappropriate study designs (letters, editorials, or case reports), or ineligible populations or interventions (e.g., corticosteroid-containing injections or paediatric patients) (n=33). One publication had been retracted (n=1). Seven studies directly addressed the clinical question and were included.
Relevant Paper(s):
| Study Title | Patient Group | Study type (level of evidence) | Outcomes | Key results | Study Weaknesses |
|---|---|---|---|---|---|
| A Randomized, Sham-Controlled Trial of Greater Occipital Nerve Block for Acute Migraine in the Emergency Department Friedman BW, Mohamed S, Robbins MS, et al. 2018 United States | Adults with acute migraine refractory to IV metoclopramide in the ED; n=28 (13 GONB, 15 sham) | Sham-controlled RCT of 28 ED patients with migraine refractory to IV metoclopramide, showing 31% headache freedom with GONB vs. 0% with sham at 30 minutes (p=0.035) | Primary: headache freedom at 30 min. Secondary: pain improvement at 1 hour | Headache freedom at 30 min: 31% GONB vs. 0% sham (95% CI 6–56, p=0.035). Trial stopped early (36% of target enrolment) | • Stopped early at only 36% of target enrollment (28 of 78 patients), severely limiting statistical power and precision • Very small sample (n=13 GONB, n=15 sham) • Studied GONB only as a rescue after metoclopramide failure, limiting generalizability to first-line use |
| A Randomized, Double-Dummy, Emergency Department-Based Study of Greater Occipital Nerve Block With Bupivacaine vs Intravenous Metoclopramide for Treatment of Migraine Friedman BW, Irizarry E, Williams A, et al. 2020 United States | Adults with moderate-severe acute migraine in the ED; n=99 (51 GONB, 48 metoclopramide) | Double-dummy non-inferiority RCT of 99 ED patients comparing first-line GONB to IV metoclopramide, which failed to establish non-inferiority (difference −1.1, margin −1.3). Subgroup analysis showed experienced injectors achieved substantially better outcomes (6.9 vs. 4.1 pain improvement) | Primary: mean pain improvement (0-10) at 1 hour. Secondary: sustained headache relief at 48 hours | Mean pain improvement: GONB 5.0 (95% CI 4.1–5.8) vs. metoclopramide 6.1 (95% CI 5.2–6.9); difference −1.1 (95% CI −2.3 to 0.1). Non-inferiority not established (margin −1.3). Experienced injectors: 6.9 vs. inexperienced: 4.1 (difference 2.8; 95% CI 1.1–4.5) | • Stopped early (99 of planned 170 patients enrolled), again limiting precision • Non-inferiority not established; the primary endpoint failed • Injectors were largely inexperienced (median of 2 prior GONBs), which the study's own subgroup analysis showed substantially degraded outcomes (4.1 vs. 6.9 pain improvement) |
| Greater Occipital and Supraorbital Nerve Blockade for the Acute Treatment of Migraine: A Double-Blind, Randomized, Placebo-Controlled Study Hokenek NM, Ozer D, Yılmaz E, et al 2021 Turkey | Adults with acute migraine in the ED; n=128 (4 arms: GONB, SONB, combined, sham; ~32 per arm) | Four-arm sham-controlled RCT of 128 ED patients comparing GONB, supraorbital nerve block, combined block, and sham, with all active arms significantly superior to sham at 120 minutes. | Primary: VAS change at 120 minutes | VAS reduction: GONB −54.1, SONB −42.0, combined −59.3, sham −9.9. All active arms vs. sham (p=0.001). GONB superior to SONB (p=0.001) | • Excluded chronic migraine and medication overuse headache, limiting applicability to common ED presentations • Relatively short follow-up (120 minutes), with no assessment of sustained relief or headache recurrence • Used lidocaine 1% rather than bupivacaine, making direct comparison with the other trials difficult |
| The effectiveness of greater occipital nerve blockade in treating acute migraine-related headaches in emergency departments Korucu O, Dagar S, Çorbacıoğlu ŞK, Emektar E, Cevik Y 2018 Turkey | Adults with acute migraine in the ED; n=60 (20 per arm: GONB, IV dexketoprofen + metoclopramide, sham) | Three-arm placebo-controlled RCT of 60 ED patients comparing GONB, IV dexketoprofen/metoclopramide, and sham at 45 minutes | Primary: pain score change at 5, 15, 30, 45 minutes | Median pain reduction at 45 min: GONB 7 (IQR 6–9) vs. IV treatment 7 (IQR 5–8) vs. sham 5 (IQR 2.25–7). GONB and IV both superior to sham (p=0.016 and p=0.03). No difference between GONB and IV (p=0.39) | • Very small sample (n=20 per arm) • Short follow-up (45 minutes only) • Single-center study with limited demographic reporting |
| Influence of greater occipital nerve block on the relief of acute migraine: A meta-analysis Li W, Tang L. 2024 China | Pooled analysis of 4 RCTs, 224 patients with acute migraine | Meta-analysis pooling 4 RCTs (224 patients). | Pain scores at 0–15, 30, and 45–60 min; sustained headache relief; rescue medication | GONB significantly reduced pain at 30 min (MD −1.95; 95% CI −2.61 to −1.29; p=0.00001) and 45–60 min (MD −2.31; 95% CI −3.08 to −1.53; p=0.00001). No significant effect at 0–15 min, on sustained relief, or rescue medication | • Pooled only 4 RCTs with 224 total patients, which is a small evidence base for meta-analysis • Significant heterogeneity in comparators, local anesthetic agents, injection volumes, and outcome timepoints • Found no significant effect on sustained headache relief or rescue medication use, limiting conclusions about durability |
| Greater Occipital Nerve Block for Migraine: An Umbrella Review and Independent Systematic Review and Meta-Analysis Atraszkiewicz D, Ünal E, Bassett P, Morell-Ducos F, Bahra A 2025 United Kingdom | 16 RCTs (930 pts) qualitative; 7 RCTs (401 pts) quantitative. Acute migraine subgroup: LA-only GONB vs. sham/placebo | Umbrella review + independent SR and meta-analysis of 16 RCTs (930 pts) qualitative; 7 RCTs (401 pts). Acute migraine subgroup: LA-only GONB vs. sham/placebo | Headache severity at 30 min (acute); headache severity and monthly headache days at 1 month (chronic); adverse events | Acute migraine (moderate certainty): LA GONB reduces headache severity at 30 min (MD −2.08; p0.001). All 9 prior systematic reviews had significant limitations. No serious adverse effects across all trials | • Identified that all 9 prior systematic reviews had significant methodological limitations, raising concerns about the reliability of the cumulative evidence base • Only 7 of 16 identified RCTs could be included in quantitative synthesis • Concluded that sustained benefits remain unclear and called for more homogeneous RCTs |
| 2025 Guideline Update to Acute Treatment of Migraine for Adults in the Emergency Department: The American Headache Society Evidence Assessment of Parenteral Pharmacotherapies Robblee J, Minen MT, Friedman BW, Cortel-LeBlanc MA, Cortel-LeBlanc A, Orr SL. 2025 United States / Canada | Adults with acute migraine in the ED; evidence synthesis of 3 class I and 1 class II RCTs | Systematic review, meta-analysis, and practice guideline synthesizing 3 class I and 1 class II RCTs | Pain reduction, headache freedom, rescue medication, adverse events | GONB received Level A recommendation ("must offer"). Three positive class I studies; GONB highly likely effective and safe. Provider experience may improve response. No severe adverse effects. | • As a guideline, its strength is entirely dependent on the underlying trials, which as noted above are uniformly small and heterogeneous • The Level A recommendation is based on only 3 class I and 1 class II studies, a narrow evidence base for the highest recommendation level • Provider experience caveat is acknowledged but no minimum training recommended |
Author Commentary:
The body of evidence supports the use of GONB with local anaesthetic as an effective intervention for acute migraine in the emergency department. Four RCTs directly address the PICO question, and three of them, Friedman et al. (2018), Hokenek et al. (2021), and Korucu et al. (2018), demonstrated statistically significant and clinically meaningful reductions in pain scores when GONB was compared to a sham injection. Hokenek et al. additionally showed GONB to be superior to supraorbital nerve block alone (p=0.001), and Korucu et al. found GONB equivalent to IV dexketoprofen plus metoclopramide (p=0.39). Notably, Friedman et al. (2020) was a double-dummy non-inferiority trial found that GONB failed to establish that GONB was non-inferior to metoclopramide.
There is notable heterogeneity across the included studies in terms of population, comparator, local anaesthetic agent and concentration, injection volume, and follow-up duration. Bupivacaine (0.5%) was used in three studies (both Friedman trials and Korucu et al.) and lidocaine (1%) in one (Hokenek et al.), reflecting the variation in GONB protocols seen in clinical practice. Of clinical relevance is the finding from Friedman et al. (2020) that provider experience substantially influenced outcomes: injectors who had performed fewer than seven prior GONBs achieved a mean pain improvement of 4.1 points, compared to 6.9 points in those with greater experience, a difference of 2.8 points (95% CI 1.1 to 4.5). This finding suggests that GONB efficacy in the ED may depend heavily on institutional training and procedural familiarity, and offers a plausible explanation for the negative primary result in the same trial, where the median number of prior GONBs performed by injectors was just two.
Li and Tang (2024) pooled four RCTs comprising 224 patients and found that GONB significantly reduced pain at 30 minutes (MD −1.95; 95% CI −2.61 to −1.29) and at 45 to 60 minutes (MD −2.31; 95% CI −3.08 to −1.53), but found no significant effect on sustained headache relief or rescue medication use. Atraszkiewicz et al. (2025) conducted an umbrella review identifying nine prior systematic reviews, all carrying significant limitations or methodological errors despite being collectively cited 256 times, and performed an independent meta-analysis yielding a similar pooled estimate (MD −2.08 at 30 minutes; p<0.001) with moderate certainty. Importantly, their review concluded that the sustained benefits of GONB remain unclear and that RCTs with more homogeneous methodology are still needed.
The 2025 American Headache Society guideline update represents the most authoritative synthesis available, classifying GONB as Level A, meaning it is a treatment that must be offered, based on three positive class I studies and one class II study, with no severe adverse effects reported across all included trials.
There is notable heterogeneity across the included studies in terms of population, comparator, local anaesthetic agent and concentration, injection volume, and follow-up duration. Bupivacaine (0.5%) was used in three studies (both Friedman trials and Korucu et al.) and lidocaine (1%) in one (Hokenek et al.), reflecting the variation in GONB protocols seen in clinical practice. Of clinical relevance is the finding from Friedman et al. (2020) that provider experience substantially influenced outcomes: injectors who had performed fewer than seven prior GONBs achieved a mean pain improvement of 4.1 points, compared to 6.9 points in those with greater experience, a difference of 2.8 points (95% CI 1.1 to 4.5). This finding suggests that GONB efficacy in the ED may depend heavily on institutional training and procedural familiarity, and offers a plausible explanation for the negative primary result in the same trial, where the median number of prior GONBs performed by injectors was just two.
Li and Tang (2024) pooled four RCTs comprising 224 patients and found that GONB significantly reduced pain at 30 minutes (MD −1.95; 95% CI −2.61 to −1.29) and at 45 to 60 minutes (MD −2.31; 95% CI −3.08 to −1.53), but found no significant effect on sustained headache relief or rescue medication use. Atraszkiewicz et al. (2025) conducted an umbrella review identifying nine prior systematic reviews, all carrying significant limitations or methodological errors despite being collectively cited 256 times, and performed an independent meta-analysis yielding a similar pooled estimate (MD −2.08 at 30 minutes; p<0.001) with moderate certainty. Importantly, their review concluded that the sustained benefits of GONB remain unclear and that RCTs with more homogeneous methodology are still needed.
The 2025 American Headache Society guideline update represents the most authoritative synthesis available, classifying GONB as Level A, meaning it is a treatment that must be offered, based on three positive class I studies and one class II study, with no severe adverse effects reported across all included trials.
Bottom Line:
The current evidence suggests GONB with local anaesthetic is a safe, rapidly effective, and well-tolerated intervention that should be incorporated into a multimodal treatment strategy for acute migraine in the ED. It is superior to sham injection and appears equivalent to standard IV pharmacotherapy in experienced hands. It is a particularly attractive adjunct for patients who are unable or unwilling to receive IV medications, those with contraindications to standard pharmacotherapy, or those in whom first-line treatments have already failed.
Level of Evidence:
Level 1: Recent well-done systematic review was considered or a study of high quality is available
References:
- Friedman BW, Mohamed S, Robbins MS, et al.. A Randomized, Sham-Controlled Trial of Greater Occipital Nerve Block for Acute Migraine in the Emergency Department
- Friedman BW, Irizarry E, Williams A, et al.. A Randomized, Double-Dummy, Emergency Department-Based Study of Greater Occipital Nerve Block With Bupivacaine vs Intravenous Metoclopramide for Treatment of Migraine
- Hokenek NM, Ozer D, Yılmaz E, et al. Greater Occipital and Supraorbital Nerve Blockade for the Acute Treatment of Migraine: A Double-Blind, Randomized, Placebo-Controlled Study
- Korucu O, Dagar S, Çorbacıoğlu ŞK, Emektar E, Cevik Y. The effectiveness of greater occipital nerve blockade in treating acute migraine-related headaches in emergency departments
- Li W, Tang L.. Influence of greater occipital nerve block on the relief of acute migraine: A meta-analysis
- Atraszkiewicz D, Ünal E, Bassett P, Morell-Ducos F, Bahra A. Greater Occipital Nerve Block for Migraine: An Umbrella Review and Independent Systematic Review and Meta-Analysis
- Robblee J, Minen MT, Friedman BW, Cortel-LeBlanc MA, Cortel-LeBlanc A, Orr SL.. 2025 Guideline Update to Acute Treatment of Migraine for Adults in the Emergency Department: The American Headache Society Evidence Assessment of Parenteral Pharmacotherapies