Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Acad Emerg Med. Author manuscript; available in PMC 2013 October 1.
Published in final edited form as:
PMCID: PMC3473153

A Randomized Controlled Trial of a Comprehensive Migraine Intervention Prior to Discharge From an Emergency Department

Monitoring Editor: Steven B. Bird, MD.



Patients who use an emergency department (ED) for acute migraine headaches have higher migraine disability scores and lower socioeconomic status and are unlikely to have used a migraine-specific medication prior to presentation to the ED. The objective was to determine if a comprehensive migraine intervention, delivered just prior to ED discharge, could improve migraine impact scores 1 month after the ED visit.


This was a randomized controlled trial of a comprehensive migraine intervention versus typical care among patients who presented to an ED for management of acute migraine. At the time of discharge, for patients randomized to comprehensive care, the authors’ protocol reinforced their diagnosis, shared a migraine education presentation from the National Library of Medicine, provided them with six tablets of sumatriptan 100 mg and 14 tablets of naproxen 500 mg, and if they wished, provided them with an expedited free appointment to our institution's headache clinic. Patients randomized to typical care received the care their attending emergency physicians (EPs) felt was appropriate. The primary outcome was a between-group comparison of the Headache Impact Test (HIT-6) score, a validated headache assessment instrument, 1 month after ED discharge. Secondary outcomes included an assessment of satisfaction with headache care and frequency of use of migraine-specific medication within that 1-month period.


Over a 19-month period, 50 migraine patients were enrolled. One-month follow-up was successfully obtained in 92% of patients. Baseline characteristics were comparable. One-month HIT-6 scores in the two groups were nearly identical (59 vs. 56, 95% confidence interval [CI] for difference of 3 = –5 to 11), as was dissatisfaction with overall headache care (17% vs. 18%, 95% CI for difference of 1% = –22% to 24%). Patients randomized to the comprehensive intervention were more likely to be using triptans or migraine-specific therapy (43% vs. 0%, 95% CI for difference of 43% = 20 to 63%) 1 month later.


A comprehensive migraine intervention, when compared to typical care, did not improve HIT-6 scores (a validated measure of the effect of migraine on one's daily life) 1 month after ED discharge. Future work is needed to define a migraine intervention that is practical and useful in an ED, where many underserved patients, of necessity, present for care.

Migraine affects 12% of the general U.S. population.1 Work absenteeism and decreased productivity at work due to this illness result in $13 billion in annual costs to U.S. businesses.2 Migraine is underdiagnosed (only 56% of migraine patients have received this diagnosis) and undertreated (49% use only over-the-counter medications to treat their headaches). Few eligible patients use preventative medications.1 Remaining undiagnosed, and therefore inadequately treated, is associated with low socioeconomic status.3

Effective migraine treatment requires accurate diagnosis, patient education, and initiation of appropriate acute and preventive medications. Diagnosis and educational interventions can improve migraine disability scores as part of a comprehensive program of care.4 Medications have demonstrated efficacy at decreasing pain and functional disability. These include triptans, nonsteroidals, and dopamine receptor blockers such as metoclopramide.5,6 Various preventive medications decrease the number of headache days suffered by patients with frequent migraine.7 Incorporating these various components of care, while addressing comorbidities and excluding organic causes of headache, represents state-of-the-art migraine treatment.

Patients who use the emergency department (ED) for treatment of severe migraine are socioeconomically depressed, have difficulty accessing primary care, and have more severe underlying disease than those who do not use the ED.8 As many as 50% of migraine patients who use the ED take no medication, not even acetaminophen, before presenting to the ED.9 Use of sumatriptan or other triptans is reported infrequently by patients who use the ED for treatment of acute migraine (B. W. Friedman, 2004 and 2006, unpublished data). However, it has been shown that prescribing sumatriptan can result in substantial improvements in overall health functioning scores, decreased absenteeism from work, and fewer health care visits.10

We conducted a randomized, controlled study to determine if a comprehensive ED-based migraine treatment protocol, which incorporated many of the features discussed above, could improve headache impact scores 1 month after an ED visit for acute headache. Our specific hypothesis was as follows: a protocol designed to diagnose, educate, treat, and refer underserved migraine patients at the time of their discharge from the ED would improve 1-month Headache Impact Test (HIT-6) scores (a validated measure of the effect of migraine on one's daily life) significantly more than typical ED discharge care.


Study Design

This was a randomized controlled trial to test the hypothesis that a program of ED-based comprehensive migraine care could improve headache outcomes 1 month after ED discharge when compared to typical ED therapy. We registered the trial at (NCT01071317). The Montefiore Medical Center institutional review board approved the trial and provided ethics oversight.

Study Setting and Population

We performed this study in the ED of Montefiore Medical Center, an urban ED serving the communities in and around the Bronx, New York. The ED receives over 90,000 adult visits annually. The research staff of the Montefiore ED are salaried, trained, technician-level bilingual (English and Spanish) research associates, who execute research studies under the supervision of the principal investigators. These research associates staff the Montefiore ED around the clock.

Patients with migraine were eligible for this study if they presented to the ED for treatment of acute headaches. Specific migraine subtypes eligible for enrollment included migraine with or without aura, as defined by the International Classification of Headache Disorders. To participate, patients were required to have baseline Migraine Disability Assessment (MIDAS) scores of 5 or more, signifying at least mild migraine-associated disability. To focus our intervention on underserved migraineurs, patients were excluded if they had current prescriptions for any triptan or migraine-preventive therapy or if they responded affirmatively to the question, “Do you have a doctor who helps you take care of your headaches?” We also excluded patients from participation for the following reasons: daily or near-daily headache or analgesic medication use, pregnant or breast-feeding, unavailable for 1-month follow-up, admitted to the hospital, previous enrollment in the study, or pain not currently severe enough to require parenteral medication, as determined by the clinical attending emergency physician (EP). After reviewing study eligibility criteria , research associates, supervised by the principal investigator, explained the study to eligible patients and then obtained their written consent to participate as research subjects.

Study Protocol

The intervention took place at two points: in person, at the time of discharge from the ED; and by telephone, 1 week after ED discharge (Figure 1). The intervention consisted of four components: reinforcement of diagnosis, migraine education, prescription of appropriate acute migraine medication, and referral to specialty care.

Figure 1
Protocol outline. HIT-6 = Headache Impact Test.

Reinforcement of Diagnosis

Research personnel explained the following: “You have a migraine headache. Migraine is usually described as a one-sided [pointed to left or ride side of head] throbbing headache, with nausea, vomiting, and sensitivity to bright lights and loud noises. Migraine sometimes comes with an aura, which means changes in your vision like shapes, shadows, or lights, or odd sensations like tingling or numbness. The questions we asked you earlier allowed us to categorize your headache as a migraine.”


We used patient educational material provided online by the National Library of Medicine and National Institutes of Health. This information, part of the Medline Plus patient tutorials, is available at Patients were given an opportunity to review the online slide show presented at this website after their pain was relieved. Each patient received a printed copy of the tutorial to take home. Research associates then reviewed highlights of the tutorial with patients. Specific points stressed included: 1) awareness of migraine triggers, 2) the range of acute treatments available, 3) the availability and role of preventive medications, and 4) how to access specialist migraine care and migraine resources.


Six tablets of sumatriptan 100 mg and 14 tablets of naproxen 500 mg were distributed to each patient at the time of ED discharge. If a participant was allergic to or had a contraindication to one of these medications, metoclopramide 20 mg was substituted for that medication.


Each patient was contacted by telephone 1 week after ED discharge and offered an expedited free appointment at the Montefiore headache clinic.

Typical Care

Patients randomized to this group received standard ED discharge care, as determined by the attending EP, and were referred to a primary care provider. Health care providers caring for patients randomized to typical care were informed that the patient was enrolled in a study examining outcomes after treatment of migraine in the ED and that they were to treat the patient as they saw fit.

Randomization and Blinding of Outcome Assessment

Group assignment was determined by random allocation, using a random number generator ( After a patient consented to participate in the study, research associates logged into the website and received the patient's assignment. Because of the nature of the intervention, it was not possible to blind patients randomized to the comprehensive intervention arm to allocation. However, the following efforts were made to blunt this limitation: 1) in the ED, all patients, regardless of assignment, received acute migraine therapy per protocol; 2) all patients received follow-up phone calls 1 week after ED discharge; and 3) most importantly, the 1-month outcome assessor was blinded to assignment.

Outcome Measures

The primary outcome for this trial was the HIT-6, a six-item headache impact test that assesses the pain and functional effect of migraine on an individual's daily life. This instrument closely tracks lengthier migraine and general health care-associated quality-of-life instruments and proves useful over the full range of migraine severity.11,12 Scores on this instrument range from 36 to 78, with higher scores indicating more substantial illness. The primary outcome assessment occurred 1 month after ED discharge. Scores were compared between the intervention group and the typical care group.

Secondary outcomes included an assessment of health care resources utilized, including ED visits and medications taken, and an assessment of the patient's satisfaction with the headache care they received (unsatisfied, mostly satisfied, completely satisfied). We also assessed the patients’ confidence in managing their migraines by having them describe themselves as uncomfortable, somewhat comfortable, or very comfortable with their ability to manage their headaches. We surveyed what medications the patients were currently using for their migraines and categorized the following as migraine-specific medication: triptans, dihydroergotamine, metoclopramide, topiramate, beta-blockers, and tricyclic antidepressants. Nonsteroidal anti-inflammatory drugs, opioids, and butalbital-containing compounds were not considered migraine-specific medication.

Data Analysis

We compared the HIT-6 scores between the intervention group versus the control group using the Student's t-test. Statistical significance for between-group differences is reported with 95% confidence intervals (CIs). If the HIT-6 scores were not normally distributed, the Mann-Whitney U-test was to be used in place of the Student's t-test. Secondary outcomes were compared statistically using the chi-square test and are also reported with 95% CI.

The sample size calculation was based on a clinically relevant effect size identified in a study of patients with chronic daily headache: the change in the HIT-6 associated with patients feeling at least “somewhat better” over time was 2.5.13 Using standard deviations (SDs) reported in Coeytaux et al.,13 a two-sided alpha of 0.05, and 80% power, we determined the need for 23 subjects in each group.


Enrollment commenced in February of 2010 and continued for 19 months. Please see the CONSORT flow diagram (Figure 2). Baseline characteristics were comparable between the groups (Table 1).

Figure 2
CONSORT flow diagram. *One patient was excluded post randomization because a late decision was made not to discharge the patient. This patient was not followed beyond the ED visit. Other exclusion criteria include admitted to hospital (5); pregnant (16); ...
Table 1
Baseline Characteristics

The HIT-6 scores 1 month after ED discharge were comparable (Table 2). Results did not change if the Mann-Whitney U-test was used in place of the Student's t-test. Similarly, patients in both arms were comparably likely to return to an ED (Table 2). Patients randomized to the intervention were more likely to be using a migraine-specific medication 1 month (Table 2) and 1 week (Table 3) after ED discharge, but this did not translate into greater satisfaction with headache care, greater confidence in their own ability to manage headaches, or fewer days with headache.

Table 2
One-month Follow-up
Table 3
One-week Follow-up

Of the 25 patients randomized to comprehensive intervention, 17 took the time in the ED to view the online educational presentation. The median time to view the presentation was 15 minutes (interquartile range [IQR] = 10 to 20 minutes). Eleven patients (65%) described it as very helpful, while six described it as somewhat helpful. Six patients requested to view more online presentations. Four patients randomized to the comprehensive intervention accepted a headache clinic appointment. At the headache clinic, these patients were given the following diagnoses: 1) migraine with and without aura; 2) hemicrania continuum with and without migraine and medication overuse headache; 3) chronic daily headache and migraine; and 4) migraine with aura, chronic migraine, depression, and facial pain.


In this randomized controlled trial of a comprehensive migraine intervention at the time of discharge from the ED, we were unable to improve outcomes in patients who received the intervention compared to those who received typical care. Although these results may be biased by the unconcealed allocation of the comprehensive intervention arm, we could find no way to avoid this feature, which seemed inherent to the study design.

It is clear that patients who use the ED for management of migraine are in need of appropriate care: many take no medication, not even acetaminophen, before presenting to an ED for treatment.14 It is also clear that use of an ED for management of headache is associated with uninsured status, lower socioeconomic status, and worse underlying migraine scores.15 Therefore, one might reasonably speculate that this group would be likely to benefit from a minimal intervention. However, in the execution of the study, we identified several unanticipated challenges that may have impeded the demonstration of benefit to our target population.

The first challenge is identifying those migraineurs most likely to benefit. In this study, we limited enrollment to those patients with baseline MIDAS scores of at least five because we believed that a benefit from the intervention would be easiest to demonstrate in migraineurs with more than minimal baseline migraine-related disability. However, it is by no means clear that higher MIDAS scores reported during a baseline ED visit signify persistently elevated MIDAS scores. In previous work, we demonstrated no association between MIDAS scores at an index ED visit and at 3 months.16 The rationale for this may be that circumstances that lead to the ED visit contribute to a transient spike in the patient's migraine related disability, which will then regress to the patient's usual level over the successive months. The only predictor of poor 3-month outcomes in our previous ED-based cohort was Medicaid insurance (vs. private insurance or no insurance).16 Going forward, we need to understand with greater clarity which ED patients are most likely to benefit from a comprehensive migraine intervention.

A second challenge is to measure the benefit. Substantially more patients in the comprehensive intervention were taking migraine-specific medication at 1 week and 1 month. Similarly, substantially more patients in the comprehensive intervention at both time points were very comfortable managing their headaches. Why was this not reflected in the HIT-6 score? Knowledge accrued during the intervention may pay dividends months to years down the road. A 1-month follow-up may not have been sufficiently removed from the event for the full effect of the intervention to have taken effect. Alternatively, the problem may be in the instrument. HIT-6 scores may take longer than 1 month to equilibrate at a stable level that reflects the treated migraine state.

A third challenge is to create a one-size-fits-all intervention, which can then be standardized. Some patients may best be served by a discussion about triggers. Other may do well with a preventive agent or by combining acute therapeutics. A dedicated physician can meticulously apply a customized treatment plan to a specific patient. A standardized protocol cannot be expected to be as effective. The goal should be to develop a protocol that is beneficial enough for most patients and one that is better than a typical physician, who may not be attuned to the subtleties of headache medicine and who may not have the requisite time to dedicate to the individual patient.

A fourth challenge is dealing with the complexity of headache diagnosis without an informed health care provider guiding every step. Some of our patients made it to the specialty headache clinic, where they typically were given several headache diagnoses. Diagnosing multiple primary headaches is beyond the capacity of any generalizable protocol.

A final challenge is educating a groggy brain. Our educational intervention was delivered shortly after headache relief was obtained, and often after the patient received a centrally acting parenteral medication. Still, all patients found it helpful, and most found it very helpful. Future work will be necessary to test retained knowledge.

Because we are still at the point of developing an efficacious intervention we did not yet focus on real-world barriers, namely, how to deliver this intervention without the research staff available in our ED and the headache resources available at our medical center. Standardizing treatments and using online educational interventions goes a long way toward creating a generalizable intervention. Then referral can be used only for the neediest patients. Future endeavors should continue to develop a generalizable intervention that could potentially make a difference for the millions of underserved migraineurs who use EDs annually.


Limitations of this work include unconcealed allocation, which we found methodologically impossible to avoid; a small sample size, which, although it was based on published data, underestimated the variability in the 1-month HIT-6 scores; and a slow rate of recruitment, which was influenced mostly by our stringent entry criteria. It is also possible that the typical care arm was “contaminated” by the study protocol, in that physicians or research associates altered their behavior. However, we think that this is less likely because no patients in the typical care arm received migraine-specific medication. It is not yet clear why emergency clinicians prescribed triptans only infrequently; this may be related to medication cost, concern about adverse medication effects, or skepticism about sumatriptan's relative efficacy.17


A comprehensive migraine intervention, when compared to typical care, did not improve Headache Impact Test-6 scores (a validated measure of the effect of migraine on one's daily life) 1 month after ED discharge. Future work is needed to define a migraine intervention that is practical and useful in an ED, where many underserved patients, of necessity, present for care.


Some of this research was conducted with Dr. Friedman's K23 Career Develoment Award from the National Institute of Neurological Disorders and Stroke 1K23NS051409.


The authors have no relevant financial information or potential conflicts of interest to disclose.


1. Diamond S, Bigal ME, Silberstein S, Loder E, Reed M, Lipton RB. Patterns of diagnosis and acute and preventive treatment for migraine in the United States: results from the American Migraine Prevalence and Prevention study. Headache. 2007;47:355–63. [PubMed]
2. Hu XH, Markson LE, Lipton RB, Stewart WF, Berger ML. Burden of migraine in the United States: disability and economic costs. Arch Intern Med. 1999;159:813–8. [PubMed]
3. Lipton RB, Diamond S, Reed M, Diamond mL, Stewart WF. Migraine diagnosis and treatment: results from the American Migraine Study II. Headache. 2001;41:638–45. [PubMed]
4. Rothrock JF, Parada VA, Sims C, Key K, Walters NS, Zweifler RM. The impact of intensive patient education on clinical outcome in a clinic-based migraine population. Headache. 2006;46:726–31. [PubMed]
5. Colman I, Brown MD, Innes GD, Grafstein E, Roberts TE, Rowe BH. Parenteral metoclopramide for acute migraine: meta-analysis of randomised controlled trials. BMJ. 2004;329:1369–73. [PMC free article] [PubMed]
6. Goadsby PJ, Lipton RB, Ferrari MD. Migraine--current understanding and treatment. N Engl J Med. 2002;346:257–70. [PubMed]
7. Dodick DW, Silberstein SD. Migraine prevention. Pract Neurol. 2007;7:383–93. [PubMed]
8. Friedman BW, Reed ML, Serrano D, Diamond M, Lipton RB. Frequency of emergency department or urgent care use: results from the American migraine prevalence and prevention study [abstract]. Headache. 2007;47:745–6.
9. Friedman BW, Greenwald P, Bania TC, et al. Randomized trial of IV dexamethasone for acute migraine in the emergency department. Neurology. 2007;69:2038–44. [PubMed]
10. Lofland JH, Johnson NE, Batenhorst AS, Nash DB. Changes in resource use and outcomes for patients with migraine treated with sumatriptan: a managed care perspective. Arch Intern Med. 1999;159:857–63. [PubMed]
11. Lipton RB. How useful is the HIT-6 for measuring headache-related disability? Nat Clin Pract Neurol. 2006;2:70–1. [PubMed]
12. Nachit-Ouinekh F, Dartigues JF, Henry P, et al. Use of the headache impact test (HIT-6) in general practice: relationship with quality of life and severity. Eur J Neurol. 2005;12:189–93. [PubMed]
13. Coeytaux RR, Kaufman JS, Chao R, Mann JD, Devellis RF. Four methods of estimating the minimal important difference score were compared to establish a clinically significant change in Headache Impact Test. J Clin Epidemiol. 2006;59:374–80. [PubMed]
14. Salomone JA, 3rd, Thomas RW, Althoff JR, Watson WA. An evaluation of the role of the ED in the management of migraine headaches. Am J Emerg Med. 1994;12:134–7. [PubMed]
15. Friedman BW, Serrano D, Reed M, Diamond M, Lipton RB. Use of the emergency department for severe headache. A population-based study. Headache. 2009;49:21–30. [PMC free article] [PubMed]
16. Friedman BW, Hochberg ML, Esses D, et al. Recurrence of primary headache disorders after emergency department discharge: frequency and predictors of poor pain and functional outcomes. Ann Emerg Med. 2008;52:696–704. [PubMed]
17. Evans RW, Friedman BW. Headache in the emergency department. Headache. 2011;51:1276–8. [PubMed]