| Medication, Dose, Route of Administration |
Summary of Evidence |
Conclusion About Efficacy |
Adverse Medication Effects |
Principles of Medication Action |
Recommendation |
| QUESTION 1 Which injectable medications should be considered first-line treatment for adults who present to an ED with acute migraine? |
| Acetaminophen 1 gm IV |
Class 1: No difference vs placebo (n = 60)
Class 2: No difference vs dexketoprofen (n = 200)
Class 3: No worse than rizatriptan 5 mg PO (n = 148) |
Possibly effective |
No serious or frequent adverse events |
Well studied in a variety of pain disorders |
May offer |
| Acetylsalicylic acid 0.5–1.8 gm IV |
Class 1: None
Class 2: Substantial benefit vs placebo (n = 40)
Benefit vs placebo, inferior to SC sumatriptan (n = 275, 3 arms)
No difference vs valproic acid (n = 40)
Class 3: Benefit vs ergotamine (n = 56, crossover) |
Likely effective |
No serious or frequent adverse events. Better tolerated than sumatriptan |
Well studied in a variety of pain conditions |
May offer |
| Chlorpromazine 0.1–25 mg IV |
Class 1: None
Class 2: Benefit vs placebo (n = 60)
Class 3: No difference vs metoclopramide
Class 3: No difference vs ketorolac |
Possibly effective |
Postural hypotension and drowsiness common |
None considered |
May offer |
| Dexamethasone 8–16 mg IV |
Class 1: No difference vs placebo (n = 205)
Class 2: Benefit vs morphine (n = 190)
Class 3: No difference vs valproic acid (n = 31)
Inferior to propofol (n = 90) |
Possibly ineffective |
No serious or frequent adverse events |
Avascular necrosis is a known serious adverse event attributable to corticosteroid use. It is probably associated with longer courses, higher doses, and cumulative doses. This complication has rarely been reported after isolated parenteral doses of corticosteroids
Loss of glycemic control may require medical management |
No recommendation |
| Dexketoprofen 50 mg IV |
Class 1: Benefit vs placebo (n = 224)
Class 2: No difference vs acetaminophen 1000 mg IV (n = 200) |
Likely effective |
No serious or frequent adverse events |
None considered |
May offer |
| Diclofenac 75 mg IM |
Class 1: None
Class 2: None
Class 3: Benefit vs placebo (n = 120)
No difference vs tramadol (n = 47)
Mixed results vs DHE (n = 34) |
Possibly effective |
No serious or frequent adverse events |
None considered |
May offer |
Dihydroergotamine
1 mg SC, IV |
Class 1: None
Class 2: None
Class 3: Mixed results vs diclofenac (n = 34)
Mixed results vs sumatriptan (n = 310) |
Possibly effective |
In class 3 study, 305 adverse events among 152 patients randomized to DHE |
None considered |
No recommendation |
| Diphenhydramine 50 mg IV |
Class 1: As adjuvant therapy, no difference vs placebo (n = 208) |
Likely ineffective |
No serious or frequent adverse events |
None considered |
May avoid. Considerations of efficacy, adverse events and alternate available therapies do not support routine use as a first line therapeutic in the ED |
| Dipyrone 1 gm IV |
Class 2: No difference vs metoclopramide (n = 27)
Benefit vs placebo (n = 134) |
Likely effective |
No serious or frequent adverse events |
Agranulocytosis occurs in fewer than 1/100,000 administrations |
May offer |
| Droperidol 2.5–8.25 mg IM |
Class 1: None
Class 2: Superior to placebo (n = 305)
Class 3: No difference vs meperidine (n = 29) |
Likely effective |
In class 2 study, akathisia (31%), asthenia (25%), somnolence (20%), and anxiety (16%) were common among those who received 2.75 mg dose |
Life threatening arrhythmias have occurred rarely after administration of this medication |
May offer |
| Ergotamine 0.5 mg SC |
Class 1: none
Class 2: none
Class 3: Inferior to acetylsalicylic acid (n = 56, crossover) |
Insufficient evidence |
In class 3 study, 13% had increased nausea and vomiting |
None considered |
No recommendation |
Haloperidol
5 mg IV |
Class 1: No difference vs metoclopramide
Class 2: none
Class 3: Superior to placebo |
Likely effective |
In class 1 study, no difference in adverse events vs metoclopramide.
In class 3 study, 80% reported adverse events including 53% with motor agitation |
None considered |
May offer |
| Hydromorphone |
Class 1: none
Class 2: none
Class 3: none |
Insufficient evidence |
|
In nonrandomized studies (class 4), opioids have been associated with increased frequency of recurrent ED visits and progression of the underlying migraine disorder |
May avoid. Considerations of efficacy, principles of medication action, and alternate available therapies do not support routine use as a first line therapeutic in the ED. |
Ketamine
0.08 mg/kg IV |
Class 1: none
Class 2: none
Class 3: Superior to placebo (n = 17, crossover) |
Insufficient evidence |
Transient insobriety and fatigue were common |
None considered |
No recommendation |
Ketorolac
30–60 mg IM, IV |
Class 1: Comparable to metoclopramide, superior to valproic acid (n = 330)
Class 3: No difference vs meperidine (n = 50)
Class 3: Inferior to meperidine (n = 31)
Class 3: No difference vs chlorpromazine |
Likely effective |
Well tolerated |
Well studied in a variety of pain disorders |
May offer |
| Lidocaine 1 mg/kg IV |
Class 1: none
Class 2: Similar to placebo (n = 25)
Class 3: none |
Possibly ineffective |
Not reported |
None considered |
May avoid |
Lysine clonixinate
200 mg IV |
Class 1: none
Class 2: none
Class 3: Superior to placebo (n = 29) |
Insufficient evidence |
In class 3 study, adverse events more common than placebo |
None considered |
No recommendation |
| Magnesium 1–2 gm IV |
Class 1: Inferior to placebo (all patients administered metoclopramide) (n = 44)
Class 2: Superior to metoclopramide + dexamethasone (n = 70)
No difference vs placebo among patients with migraine without aura. Superior to placebo among patients with migraine with aura. (n = 60)
Class 3: Superior to placebo (n = 30)
Comparable to placebo and metoclopramide (n = 113) |
Insufficient evidence |
Flushing reported in several studies |
None considered |
No recommendation |
| Meperidine 75–1.5 mg/kg IM |
Class 1: none
Class 2: none
Class 3: No difference vs droperidol (n = 29)
Superior to ketorolac 30 mg IM (n = 31)
No difference vs ketorolac 60 mg IM (n = 50) |
Possibly effective |
No substantial differences vs active comparators in class 3 studies |
In nonrandomized studies (class 4), opioids have been associated with increased frequency of recurrent ED visits and progression of the underlying migraine disorder |
No recommendation |
Metoclopramide
10–20 mg IV |
Class 1: No difference vs sumatriptan (n = 78)
No difference vs prochlorperazine (n = 77)
No difference vs ketorolac, superior to valproate (n = 330)
No difference vs haloperidol (n = 64)
Class 2: No difference vs dipyrone (n = 27)
Class 3: No difference vs chlorpromazine (n = 44)
No difference vs magnesium or placebo (n = 113)
Superior to sumatriptan (n = 124) |
Highly likely to be effective |
Akathisia occurs in a minority of patients. No substantial differences vs active comparators |
None considered |
Should offer |
| Morphine 0.1 mg/kg IV |
Class 1: none
Class 2: No clinically significant difference vs dexamethasone (n = 190)
Class 3: none |
Possibly ineffective |
Not reported in class 2 study |
In non-randomized studies (class 4), opioids have been associated with increased frequency of recurrent ED visits and progression of the underlying migraine disorder |
May avoid |
| Nalbuphine |
Class 1: none
Class 2: none
Class 3: none |
Insufficient evidence |
No data available |
None considered |
No recommendation |
| Octreotide 0.1 mg SC, IV |
Class 1: No difference vs placebo (n = 43)
Class 2: Superior to placebo (n = 29)
Inferior to prochlorperazine (n = 44)
Class 3: none |
Possibly ineffective |
Local reactions and diarrhea reported in 20% of patients in class 1 & 2 studies |
None considered |
May avoid |
Prochlorperazine
10 mg IV |
Class 1: No difference vs metoclopramide (n = 77)
Superior to sumatriptan (n = 66)
Class 2: Superior to octreotide (n = 44)
Class 3: Superior to valproate (n = 44) |
Highly likely to be effective |
Akathisia and drowsiness were common |
None considered |
Should offer |
| Promethazine |
Class 1: none
Class 2: none
Class 3: none |
Insufficient evidence |
No data available |
None considered |
No recommendation |
| Propofol 10–40 mg IV initial bolus + 10–20 mg IV boluses |
Class 1: none
Class 2: Superior to sumatriptan at 30 minutes (n = 90)
Class 3: Superior to dexamethasone up to 45 minutes (n = 90) |
Possibly effective up to 45 minutes |
In class 3 study, 44% had sedation. Oxygen desaturation occurred in 4% |
Very brief duration of action |
No recommendation |
| Sumatriptan 6 mg SC |
Class 1: Comparable to metoclopramide (n = 78)
No difference between trimethobenzamide but inadequately powered (n = 40)
Inferior to prochlorperazine (n = 66)
Superior to placebo (n = 158)
Superior to placebo (n = 639)
Superior to placebo (n = 51)
Superior to placebo (n = 577)
Class 2: Superior to placebo (n = 1104)
Superior to placebo and acetylsalicylic acid (n = 275)
Superior to placebo (n = 277)
Superior to placebo (n = 86)
Superior to placebo (n = 76)
Superior to placebo (n = 242)
Inferior to propofol (n = 90)
Superior to placebo (n = 235)
Superior to placebo (n = 266)
Class 3: Superior to placebo (n = 136)
Superior to placebo (n = 200)
Superior to placebo (n = 209)
Superior to placebo (n = 170)
Superior to placebo (n = 138)
Inferior to metoclopramide (n = 124)
Mixed outcomes vs DHE (n = 310) |
Highly likely to be effective |
In ED-based studies, adverse events in 50% of patients |
Most effective if administered very early after migraine onset |
Should offer |
| Tramadol 100 mg IM |
Class 3: No differences vs diclofenac (n = 47) |
Insufficient evidence |
13% of patients reported adverse event in class 3 study |
None considered |
No recommendation |
| Triamcinolone 40 mg SC |
Class 1: none
Class 2: none
Class 3: In study of greater occipital nerve block, no difference at 20 minutes among those randomized to triamcinolone + bupivacaine + lidocaine vs bupivacaine + lidocaine alone (n = 37) |
Insufficient evidence |
No adverse events reported |
None considered |
No recommendation |
| Trimethobenzamide 200 mg IM |
Class 1: No difference between sumatriptan but inadequately powered (n = 40) |
Insufficient evidence |
5% of patients reported adverse events in class 1 study |
None considered |
No recommendation |
| Valproic Acid 500–1000 mg IV |
Class 1: Inferior to ketorolac and metoclopramide (n = 330)
Class 2: No difference vs acetylsalicylic acid (n = 40)
Class 3: No difference vs dexamethasone (n = 31)
Inferior to prochlorperazine (n = 40) |
Possibly effective |
Minimal adverse events |
None considered |
May offer |
| QUESTION 2 Do parenteral corticosteroids prevent recurrence of migraine in adults discharged from an ED? |
| Dexamethasone 8–24 mg IV |
Class 1: Meta-analysis of 3 placebo controlled class 1 studies demonstrated benefit (n = 358)
Class 2: Benefit vs morphine (n = 190)
Class 3: No difference vs valproic acid (n = 31) |
Highly likely to be effective |
Dizziness and brief burning pain more common in dexamethasone group. |
Avascular necrosis is a known serious adverse event attributable to corticosteroid use. It is probably associated with longer courses, higher doses, and cumulative doses. This complication has rarely been reported after isolated doses of corticosteroids.
Loss of glycemic control may require medical management. |
Should offer |
