Abstract
Introduction
Migraine is a debilitating disorder which has a serious impact on the patient’s life (1). The most recent estimate of the prevalence of migraine in adults in France is 11% (2); the proportion is almost doubled if those meeting the criteria for ‘probable migraine’ (categories 1.6.1 and 1.6.2 of the 2004 International Headache Society [IHS] criteria (3)) are included (2).
Despite the establishment of recommendations in France by the Agence Nationale d’Accréditation et d’Évaluation en Santé (ANAES; the National Agency for Health Accreditation and Evaluation) for the diagnosis and treatment of migraines (4), this disorder continues to be under diagnosed (1). This is likely attributable to patients failing to consult a physician for their condition, often instead believing that they have sinusitis or tension-type headache (5). Indeed, the single major factor associated with self-awareness of migraine is current or prior medical consultation, according to the 2003 FRAMIG3 French population-based survey (6). This survey also showed that medical consultation was the single major factor associated with the use of recommended acute treatment for migraine (6). Together with the earlier FRAMIG surveys (7,8), this study highlighted the unmet need of appropriate and adequate migraine management and patient education.
However, these studies did not report the proportion of patients attending an emergency department because of their migraine. Although general practitioners are usually the first physician to be consulted (6), migraine patients may go to an emergency department in cases of attacks that are particularly severe, for initial episodes, or when their migraine is unimproved by usual treatment. About half of patients presenting with a headache were diagnosed as having a migraine (with or without aura) at a specialised headache emergency centre in Paris in a retrospective analysis of data collected at this centre from 70,000 patients seen over a period of 7 years (9). Data regarding migraine patients reporting to an emergency department in France are currently limited to data from this headache centre in Paris.
Studies in other countries suggest that migraines are under- or misdiagnosed; consequently, patients are often undertreated or receive inadequate care when they consult an emergency department (10–13). Given the lack of corresponding prospective French studies of patients with migraine presenting at a general emergency department, we conducted the ‘Enquête de la Migraine aux Urgences’ (EMU) study to establish the frequency of migraine diagnosis among patients consulting emergency departments for headaches, to determine the demographic and clinical characteristics of these patients, and to describe the treatment and follow-up they received.
Patients and methods
This was a prospective, observational, national, multicentre study of adult patients admitted to one of at least 20 general emergency departments across France during a 1-week period. For inclusion, the emergency departments had to have at least 35,000 admissions per year. No other selection criteria relating to geographic position, type of institution (academic vs non-academic hospitals), and source of funding (private vs public) were applied when approaching emergency departments to participate in the study. The study was approved by the French National Medical Council Committee and the National Data Protection Committee and conducted in accordance with Good Clinical Practice and the Declaration of Helsinki.
Patients
Patients were recruited during each 24-h period of seven consecutive days. Inclusion criteria were: age ≥18 years, consultation for headache, headache disorders, migraine or facial pain, patients able to respond to study questions and who agreed to participate in the study. Informed consent was obtained. Only patients who refused to participate were not included.
Study end-points
There were three primary end-points. First, to estimate the proportion of patients consulting for headache, headache disorders, migraine or facial pain and, among these patients, the proportion of those diagnosed with migraine; second, to characterise patients diagnosed with migraine including their demographic characteristics and migraine history; and finally, to describe the therapeutic management of patients diagnosed with migraine at emergency departments.
Other end-points included characterisation of headaches in patients presenting at an emergency department, and a description of patient management according to type of migraine and after discharge.
Data collection
Data were collected from the local register of patients consulting each emergency department (centre form and patient case report form). Time of day, reason for admission, age and sex of each patient was recorded. Student clinicians completed a questionnaire to provide demographic and disease history data and the ID migraine questionnaire (14). The ID Migraine screener three-item subset of disability, nausea and sensitivity to light is a valid, reliable, screening instrument for migraine that does not take long to complete. It has a sensitivity of 0.81 (95% CI 0.77–0.85), specificity of 0.75 (95% CI 0.64–0.84) and positive predictive value of 0.93 (95% CI 89.9–95.8) (14). The study investigators (emergency department medical doctors) completed another questionnaire with items such as: (i) diagnosis according to the latest IHS criteria (3); (ii) disease history and treatment; (iii) reason for emergency department visit; (iv) whether any further examinations were requested; and (v) treatment prescribed including any prescription at discharge. Migraine acute treatment was recorded by type as follows: non-opioid analgesics excluding non-steroidal anti-inflammatory drugs (NSAIDs; this group of drugs is simply referred to as ‘non-opioid analgesics’), NSAIDs, weak opioid agonists (
A follow-up telephone interview was conducted 6–8 weeks after admission to the emergency department to determine the patient’s rating of the quality of care received in the emergency department, their adherence to investigator instructions, and whether any episodes of migraine occurred during the follow-up period.
Patient study populations
The screened patient population included all patients consulting the emergency department aged 18 years or older; of these, those with non-trauma related headache, headache disorder, migraine or facial pain as their reason for consulting the emergency department were included in the ‘headache’ population. Of these, those with a ‘yes’ next to the question ‘is the patient suffering from a migraine’ on the investigator questionnaire were included in the ‘migraine’ population. The follow-up population included all patients in the migraine population who agreed to participate in the follow-up telephone interview.
Sample size and descriptive analyses
The inclusion of 22 centres each with 90–100 consultations per day was expected to provide, over 7 days, between 140 and 420 patients presenting with headache, assuming 1–3% of all admitted patients would have headache as previously reported in the literature (15–18). Of these, 91–273 were expected to be diagnosed with a migraine, assuming 65% of patients admitted for headache would be diagnosed with migraine (15–18). The study scientific committee agreed to an arbitrary figure of at least 35,000 admissions per year (96 admissions per day) as the requirement for inclusion of an emergency department because they estimated this would be high enough to yield patients with the desired pathology present in 0.6–2% of all patients admitted. Descriptive statistics such as mean, SD, median, minimum, maximum, and 95% confidence interval (CI) were used to summarize quantitative variables, and frequencies and the 95% CI values for normal approximation were used to summarise qualitative variables. Statistical analyses were carried out using SAS v9.1.3 on Windows XP.
Exploratory analyses
Exploratory analyses examining the concordance between diagnoses of the investigator and those derived from the ID Migraine data were conducted.
Results
Between 19 March 2007 at 9 a.m. and 26 March 2007 at 9 a.m., 17,556 patients were admitted to emergency departments at 22 centres in France. Of these, 15,835 were included in the screened population (1721 were under 18 years of age or had no age data recorded; Figure 1), of whom 10,521 had non-trauma related reasons for admission. Excluding trauma-related cases, the proportion of patients with headache, headache disorders, migraine or facial pain was 3.1% (483 of 15,835; 95% CI 2.8–3.3), representing 4.6% of patients admitted for non-trauma related reasons. Only four patients were excluded because they were not willing to participate and did not complete their questionnaires; the remainder met inclusion criteria and thus there were 479 patients in the headache population (Figure 1).
Patient selection and study populations; ED, emergency department.
Prevalence of headache and migraine in emergency departments
The proportion of patients with migraine in the screened population was 0.6% (98/15,835; 95% CI 0.5–0.7), of primary non-migraine headache 0.7% (107/15,835; 95% CI 0.6–0.8) and of secondary headache 1.7% (266/15,835; 95% CI 1.6–1.8).
Migraine diagnosis in the headache population
Migraine diagnosis in the headache population according to method of diagnosis
NA, not applicable.
More patients with a secondary headache than those with either a primary non-migraine headache or migraine received a complementary examination (62.8% vs 43.5%;
Demographics and clinical characteristics of patients with migraine
Characteristics of patients in the migraine population (
Type of migraine not known for one patient (data missing).
Among patients for whom this was not the first episode (
More patients with a first episode than those with a prior history of migraine had complementary examinations carried out for diagnosis (73.3% of 15 patients vs 35.8% of 81 patients;
Migraine treatment received in the emergency department
Figure 2 shows pharmacological migraine acute treatment given in the emergency department according to type of migraine. Non-opioid analgesics and NSAIDs were the most commonly prescribed pharmacological agents, being given in 61.2% and 42.9% of 98 patients with migraine, respectively. Triptans were not given as frequently (11.2%), but were given more often to patients with migraine without aura (10 of 77 patients) than those with migraine with aura (1 of 20 patients).
Proportion of patients in the migraine population receiving each type of pharmacological acute migraine treatment in the emergency department. NSAID, non-steroidal anti-inflammatory drug. *Excluding NSAIDs. The type of migraine was not known for one patient (data missing).
About 9% of patients did not receive any pharmacological treatment, and most patients did not receive any treatment for symptoms associated with their migraine (75% of migraine patients with aura, 72.7% without aura). Treatment for associated symptoms included anti-emetics, given to 11.7% of the 77 patients with migraine without aura and to 5% of the 20 patients with aura, and anxiolytics and/or a combination of anxiolytics and anti-emetics given to 5% of those with aura and 1.3% of those without aura.
Non-pharmacological management included advice to return to their usual practitioner or a specialist for regular medical follow-up, but less than half of the patients received this advice (40.0% of 20 patients with migraine with aura and 37.7% of 77 patients with migraine without aura). By comparison, in the subgroup of patients who were not already regularly consulting a doctor for their migraine, 6 of 12 (50%) patients with migraine with aura and 18 of 41 (43.9%) patients with migraine without aura were given this advice.
Pharmacological treatment for migraine prescribed on discharge from emergency department (migraine population)
Data are reported as percentage of patients unless stated otherwise.
Follow-up care and management of patients with migraine
The mean time to follow-up was 6.7 ± 0.8 weeks after visiting the emergency department (
Few patients could continue with usual daily living activities the same day as visiting the emergency department (8.8% of 92 patients); most did so within 1–2 days (49.5%). About one-fifth were able to resume their usual activities only after more than 1 week after their emergency department consult (26.4%). A sub-group analysis showed numerous differences in patient management between those who were able to resume their normal activities within 1 week of attending the emergency department ( Comparison of patient management in the follow-up period between patients who resumed normal activities within 1 week of the post-emergency department visit and those who did so after more than 1 week following this visit (follow-up patient population; 
Seventy percent (64/92) of patients reported experiencing further headaches in the follow-up period. Of these, 45.6% (29/64) of patients consulted a doctor, of whom 57.7% consulted their own general practitioner and 34.6% consulted a specialist. Patient compliance with recommendations to seek follow-up medical attention with their general practitioner or a specialist was about 55.6%; conversely, just over a half of those patients who were not advised to seek regular medical attention made an appointment with a doctor in the follow-up period (52.7% of 55 patients).
Exploratory analyses
Exploratory analyses of data from the headache population (excluding first–episode migraine;
Discussion
Headaches represented 3.1% of the reasons for admission to emergency departments in France, which is broadly similar to previously reported estimates of 1–2% in other countries (19–22). Of those reporting to the emergency department with headache, 21% were diagnosed with migraine in our study. Previous estimates range from 17–64% (19–22). Variations in results between epidemiological studies can be attributed to differences in study design (sampling methods, whether prospective or retrospective), criteria for migraine diagnosis, and to differences between countries in which the studies were conducted (such as access to medical care (23) or the profile of patients seeking treatment at an emergency department,
One of the strengths of our study was the use of common diagnostic criteria based on IHS criteria; thus, migraine diagnosis was standardised across centres and between investigators. However, underdiagnosis of migraine was not all together avoided in our study as reflected in Table 1, where there is a large discrepancy in the number of patients diagnosed with migraine between the ID Migraine test (
In our study, 40% (197/479) of patients in the headache population, 60% (9/15) of those with a first episode of migraine and 28% (23/81) of those with previous diagnosis of migraine had a CT scan to aid in diagnosis. Consultation with a neurologist was also relatively common in our study (one-third of patients in the headache population). Possible reasons for these additional tests and/or referrals to a specialist could be the fact that patients were experiencing an unusually severe attack, because a differential diagnosis was deemed necessary for patients reporting a first episode, or because the investigators were being more cautious in their approach because they were participating in a clinical trial.
One of our primary aims was to characterise patients who go to an emergency department because of their migraine. The larger proportion of women than men with migraine in our study, and their mean age, was broadly similar to that in previous studies (9,26). Disease duration was, however, shorter in most of our patients as compared with the general migraine population in France (45% of patients with disease duration of 1–10 years vs 40% of patients with disease duration of more than 15 years (2)). We postulate that our patients may have had less knowledge about migraine, may not yet have consulted their regular physician or a neurologist for their migraine attacks and, consequently, had poorly managed disease. They may, therefore, have presented at the emergency department because they had no specific instructions on what to do in the event of an unusually severe attack and because their current treatment was ineffective. Our data seem to support the latter suggestion because, at admission, about 90% of patients in our study reported having taken treatment for their acute attack and lack of treatment efficacy was the reason 20% of our patients consulted the emergency department. This appears to reflect the high incidence of treatment failure among the general migraine population in France (38.6% of patients (27)), which may be attributable to wide-spread self-medication with over-the-counter treatments rather than use of migraine-specific treatments (1).
In our study, most patients were given pharmacological treatment in the emergency department with non-opioid analgesics and NSAIDs. These non-specific analgesics are recommended for an acute attack only where patients previously treated with these agents experience adequate pain relief (1). If pain relief 2 h after taking an NSAID is not adequate, migraine-specific treatment with a triptan as rescue analgesic is recommended (1). If after three attacks NSAIDs are still not efficacious or tolerated, a triptan should be given in the first instance when the next migraine attack occurs (1). Triptans were not commonly given in the emergency department in our study but, when prescribed, were given principally to patients with migraine without aura (13% vs 5% of patients with aura). Moreover, a small proportion of patients were given opiates as acute treatment, yet this is not recommended (1), while ergot derivatives were not prescribed in any patient even though they are recommended alternatives to triptans in patients eligible for migraine-specific treatment. Our results suggest that many emergency department medical doctors are not aware of migraine treatment guidelines and that their choice of prescription probably depends on their preference or experience (as noted before (27)). Although French guidelines for the management of migraine (published by the ANAES (4)) do not specifically address management of the patient presenting at the emergency department, recommendations for acute migraine treatment are made (1). Given that these clinical practice guidelines are available, our study highlights the need for better education of emergency department medical doctors regarding the existence of these guidelines, and how their recommendations can be applied in daily clinical practice.
The FRAMIG population-based study highlighted the importance of continued medical consultation for migraine management (6); however, in our study, fewer than half of the patients were advised on discharge to seek regular medical follow-up (whether already receiving regular medical follow up or not). In our study, of those advised to seek follow-up, only 55% made a follow-on medical appointment with their general practitioner or a specialist. Lack of satisfaction with the first consultation does not seem to explain why patients in our study did not make a follow-on medical appointment (70% of the 92 patients in our follow-up population were satisfied with the care they received in the emergency department).
A migraine treatment is considered effective if pain relief occurs within 2 h of administration of the drug (1). In our study, only one-fifth of patients no longer had headache at the time of discharge (median time in emergency department of 4 h), and one-third of patients continued to have pain up to 2 days after their emergency department visit. Given that migraine-specific treatments, such as ergot derivatives and triptans, and treatment for associated symptoms were not given very often in our study, it is perhaps not surprising that better migraine management was not achieved in more patients. In contrast, an emergency centre in Paris solely admitting patients with headache used migraine specific treatments (
One of the limitations of our study was that acute treatment in the emergency department was not captured in enough detail: we did not record the use of specific drugs, their dosages or combinations of drugs. Also, more patients who had a migraine diagnosis had been referred to a neurologist; this could be a potential source of bias when interpreting the diagnosis data. We did not directly assess whether neurologists would be more likely to diagnose migraine or whether they were trained in IHS criteria. Future studies could address these questions and compare diagnosis of migraine in the emergency department population by neurologists with diagnosis by headache specialists. Nevertheless, the scale of our study (> 17,000 patients attended emergency departments during our study inclusion period), the geographic spread of centres across the country, and the standardised diagnostic criteria used by study investigators make a unique and valuable contribution to our understanding of emergency department practices in the management of patients presenting migraine.
Conclusions
To our knowledge, this is the first prospective, multicentre, national study estimating the frequency of patients with migraine consulting a general emergency department in France. One-fifth of patients with headache were diagnosed with a migraine. Our results show that migraine is under diagnosed and under treated. Patients often received non-specific medications, which meant they left the emergency department without symptom relief, and fewer patients than expected were given migraine-specific prophylactic treatment or advice to seek follow-up care. Our results highlight the need for improvement not only in patient management but also in patient and clinician education.
Footnotes
Acknowledgement
The authors thank Tracy Harrison of Wolters Kluwer Health Medical Communications who provided medical writing services funded by Pfizer, France.