Conflict of Interest: R. Zielman received support for conference visits from Menarini and Allergan. I.F. de Coo received support for a conference visit from Electrocore. The other authors declare that they have no conflict of interest.
Financial support: None.
Migraine is a common illness in children associated with a negative impact on the quality of life. In the Netherlands, treatment of migraine is commonly performed by general practitioners (GPs). The migraine guideline of the Dutch College of General Practitioners recommends inactivity and acetaminophen in patients with migraine who are younger than 18 years of age.
The aim of our study was to evaluate the pharmacological treatment of migraine in children by GPs before referral to the hospital. Our objective was to answer the following questions. First, are GPs inclined to prescribe medication not listed in the Dutch College of General Practitioners Guideline? Second, which clinical characteristics are associated with the use of medication not listed in this guideline?
In this retrospective cross-sectional study, prescribed medication and migraine characteristics were investigated in Dutch migraine patients (age <18 years), using hospital records and a paper-and-pencil questionnaire.
A total of 223 children were included. Medications not listed in the guideline were used in 41.3% of the patients before referral. In children younger than 12 years, the use of medication not listed in the guideline was associated with an older age, when compared with children who were treated according to the guideline. In the group of patients older than 11 years, the use of medication not listed in the guideline was associated with a longer history of migraine and a longer duration of the migraine attacks.
Medications not listed in the GPs guideline were used in a large portion of the patients younger than 18 years with migraine who were referred to secondary care.
International Classification of Headache Disorders, second edition
non-steroid anti-inflammatory drugs
Migraine is a common illness in children, with a prevalence ranging from 3% in primary school children to approximately 20% in adolescents. Migraine in children results in an average of 9 missed schooldays a year. Furthermore, the overall quality of life is lower in children with migraine compared with children without migraine, and the illness greatly affects family and caregivers.
General practitioners (GPs) play an important role in diagnosing and treating migraine. Adequate treatment of the migraine attacks is essential in children, because it improves their quality of life. The Dutch College of General Practitioners (DCGP) developed a guideline for the diagnosis and treatment of migraine to support GPs in providing optimal treatment for patients with migraine. The current guideline gives the GP less acute and prophylactic treatment options for children than for adults. For the acute treatment of migraine in patients younger than 18 years, only inactivity and acetaminophen are advised and no prophylactic treatment options are provided. Acetaminophen is not always effective and it has been reported that ibuprofen is at least twice as effective in aborting the headache during a migraine attack in children. The treatment recommendations given in GP guidelines differ between countries. For example, in a GP guideline used in the UK, more treatment options, like ibuprofen and triptans, are recommended for the treatment of migraine in patients younger than 18 years of age. Therefore, it is questionable whether the current DCGP guideline is sufficient to support the Dutch GPs in treating migraine in children.
Evaluation of the DCGP guideline for adults with migraine showed an underutilization of guideline-listed medication in the primary care of migraine patients. However, no evaluation has been published on the extent to which the DCGP guideline for the treatment of migraine in children is actually used by GPs.
The overall aim of this study was to evaluate the pharmacological treatment of migraine in children by GPs in accordance to the DCGP guideline before referral to the hospital. The following questions were addressed. First, are GPs inclined to prescribe medication not listed in the DCGP? Second, which patient characteristics are associated with the use of medication not listed in the DCGP guideline?
Study Setting and Patient Selection
This retrospective cross-sectional study was conducted at the Isala Clinics in Zwolle, a general regional hospital in the Netherlands. We selected patients younger than 18 years who were registered as having migraine from the diagnostic-treatment-combination (DTC) registration database between January 2006 and June 2011 (n = 349). The DTC is an administrative system for the intramural curative and somatic health care in the Netherlands. The included patients met the following criteria:
Younger than 18 years.
Migraine as the main reason for referral.
The symptoms fulfilled the International Classification of Headache Disorders second edition (ICHD-II) criteria of migraine without aura, migraine with aura, childhood periodic syndromes that are commonly precursors of migraine or probably migraine (ICHD-II 1.1, 1.2, 1.3, and 1.6).
Naive patients who visited a neurologist for migraine or headache for the first time.
Consultation took place at the outpatient department or headache clinic.
The patients who met the inclusion criteria were classified into a younger group (11 years of age or younger) and an older group (12 to 17 years of age). These groups were chosen because triptans are not registered in the Netherlands for patients below the age of 12.
This study met the criteria for an exemption for approval by the medical ethical committee.
Classification and Definition
Migraine was diagnosed according to the ICDH- II criteria. Most patients in this study were registered in the following categories: migraine without aura, migraine with aura, childhood periodic syndromes, and probable migraine. If patients suffered from headaches other than migraine, but migraine was the main reason for referral, the migraine was considered their primary headache.
Data were collected retrospectively. The hospital records of all patients who met our inclusion criteria were obtained. To obtain additional information from the patients who visited the regular outpatient department or headache clinic, a detailed questionnaire was presented to all included patients or their parents in December 2011. This paper-and-pencil questionnaire inquired about the characteristics of the migraine attacks and their medication status prior to referral. In January and February 2012, non-responders received 2 phone calls as a reminder to complete the questionnaire. Patients who did not return the questionnaire were not excluded from this study. Their missing data have been reported as missing.
DCGP Guideline “Headache.”
The DCGP guideline about headache contains a section on migraine in patients younger than 18 years. It recommends inactivity and acetaminophen for the acute treatment of migraine in this specific group of patients as mentioned in the Figure. If symptoms are severe or the frequency of migraine attacks is at least twice a month, referral to a neurologist or pediatrician is recommended. Prophylactic treatment should be prescribed by a neurologist or pediatrician.
SPSS for Windows version 20.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analyses and P < .05 was considered to be statistically significant. The Fisher's exact test was used for ordinal parameters and the Mann-Whitney U-test for continuous parameters. Missing data were excluded in the statistical analyses.
Patient Population and Characteristics
A total of 349 patients younger than 18 years were registered in the DTC for migraine during the 5.5-year period of this study. Of these patients, 223 met the inclusion criteria and 126 patients were excluded: 100 did not meet the ICHD-II criteria for migraine; in 22 patients, it was not the first visit to a neurologist for headache or migraine and 4 patients were admitted to the hospital or seen at the emergency ward at the time of consultation. Hospital records of all 223 patients were available. The questionnaire was returned by 152 out of the 223 included patients (68.6%). All 223 patients were included in the analyses.
Demographic characteristics and headache diagnoses are presented in Table 1. The participants age ranged from 4.3 to 17.8 years, with a mean age of 12.8 years. The younger group of patients, below 12 years, consisted of 41 females and 42 males. The older group, 12 to 17 years of age, consisted of 86 females and 54 males. A positive family history for headache, of which migraine was most prevalent, was present in 78% of the patients. A majority of the patients experienced nausea (71.7%) during migraine attacks, and approximately half (49.3%) also vomited. Photophobia and phonophobia during a migraine attack were reported in respectively 66.8% and 58.7% of the patients.
Table 1. Demographic Characteristics and Diagnosis According to the ICHD-II (n = 223)
Younger Group (<12 Years) (n = 83)
Older Group (≥12 Years) (n = 140)
All Patients (n = 223)
aSome patients were diagnosed with more than 1 subtype of migraine.
bSome patients were diagnosed with more than 1 subtype of additional headache.
ICHD-II = International Classification of Headache Disorders, second edition; SD = standard deviation.
Migraine without aura (57.0%) was most frequently diagnosed. Additional primary and secondary headaches according to the ICHD-II criteria were reported in 26 patients, medication overuse headache being most frequently reported (6.3%).
Migraine Treatment before Referral
The pharmacological treatment of the patients with migraine before referral is presented in Table 2. Some patients used both medication listed and medication not listed in the DCPG guideline. Acetaminophen was most frequently used. Before referral, non-steroidal anti-inflammatory drugs (NSAIDs) were used in 8 patients (9.6%) in the younger group and in 45 patients (32.1%) in the older group. A total of 24 patients (10.7%) used a triptan, 2 patients in the younger group, and 22 in the older group. Only 7 of the younger patients and 15 of the older group of patients had used an antiemetic before referral.
Table 2. Medication Use Before Referral
Younger Group (<12 Years) (n = 83)
Older Group (≥12 Years) (n = 140)
A patient could use multiple types of drugs at any time before consultation at the hospital.
ASA = acetyl salicylic acid.
Unknown type antiemetic
Propranolol 40 mg
Propranolol 80 mg
Prophylactic treatment had been provided to 7.2% of the patients in the younger group and 14.3% of the patients in the older group. Propranolol was the only prophylactic drug prescribed in the younger group of patients, while in the older group other prophylactic drugs had been prescribed as well.
A total of 92 patients (41.3%) used medication not listed in the DCPG guideline prior to referral of which 73 patients (52.1%) of the older group. Furthermore, 25 of these 92 patients were using more than 1 type of medication not listed in the DCPG guideline.
Migraine Characteristics and the Use of Not-Listed Medication According to the DCGP Guideline
Table 3 demonstrates the patient characteristics of those who received treatment according to the DCGP guideline and those who used medication not listed in the DCGP guideline. The migraine characteristics according to the ICHD-II criteria were not associated with medication prescription by GPs. However, other factors were significant different between listed and not-listed medication users. In the younger group, the patients using medication not listed in the DCGP guideline were older than patients using only listed medication (P < .05). In the older group, patients using medication not listed in the DCGP guideline reported a longer history of migraine (P < .01) or were having longer lasting migraine attacks (P < .01).
Table 3. Migraine Characteristics Associated With Using Medication Listed in the DCGP Guideline “Headache” at Time of Consultation With a Neurologist (n = 223)
Younger Group (<12 years)
Older Group (≥12 years)
Using Guideline Medication (n = 66)
Using Medication not Mentioned in the Guideline (n = 17)
Using Guideline Medication (n = 73)
Using Medication not Mentioned in the Guideline (n = 67)
P values depicted in bold indicate significant differences (P < .05), using Fisher's exact test or Mann-Whitney U-test appropriately; P values uncorrected for multiple testing.
This retrospective study reports on the pharmacological treatment of children, patients younger than 18 years of age, with migraine by GPs before referral to the hospital. We compared the medication use of these children with the advice as provided by the DCGP guideline. An important finding from this study is the large proportion of participants that used medication not listed in the guideline.
In the group of children younger than 12 years, the use of medication not listed in the guideline was associated with an older age, when compared with children who were treated according to the guideline. In the group of children older than 11 years, the use of medication not listed in the guideline was associated with a longer history of migraine and a longer duration of the migraine attacks.
Of all medication not listed in the DCGP guideline, NSAIDs were most frequently used to treat the headache during a migraine attack. This could be explained by the over-the-counter availability of NSAIDs in the Netherlands. A previous Dutch study demonstrated that 82% of acute pain-relieving drugs are bought over-the-counter by parents of children suffering from headache, whereas only 18% is prescribed by a physician. Moreover, it was reported that ibuprofen is twice as likely to stop the headache of migraine within 2 hours when compared with acetaminophen. At least 32% of children with migraine require stronger pain medication than ibuprofen to reduce the headache during a migraine attack. Therefore, it would be reasonable to suggest expansion of the acute medication list in the DCGP guideline to provide GPs with more primary care treatment options instead of referral.
Our findings are in contrast with the evaluation of the DCGP guideline for adults with migraine. An underutilization of listed medication according to the DCGP guideline was observed in adults with migraine before referral to a neurologist. This contrast in findings can be explained by the large difference in the DCGP guideline for migraine between children and adults. In the DCGP guideline, more pharmacological treatment options are available for adults with migraine. Adding treatment options to the DCGP guideline for children could be beneficial, as inadequate medical treatment and self-medication with over-the-counter analgesics may lead to medication overuse headache. This study demonstrates that at least 6.3% of the patients suffered from medication overuse headache at moment of referral. Another study showed that at least 9.7% of the children with migraine are using daily analgesics.[6, 12] By effectively treating migraine attacks in these patients with a different kind of acute or prophylactic medication, medication overuse headache could be prevented in these children with migraine. When medication is prescribed by a GP, information on the use of medication and side effects can be provided.
At the time of consultation, 9.4% of the patients used a triptan, which was most common in the older group. In France, a comparable frequency of triptan use (9.1%) was reported. On the contrary, a study from the UK reported triptan use in only 3.5% of the pediatric population. This difference could be explained by a difference in study population. Both the current study and the study in France reported the treatment of migraine in children by GPs at the time of referral to a hospital, whereas the study in the UK studied investigated the whole population of children with migraine treated in the primary care. Moreover, there may be a difference in severity of migraine attacks between the study populations. However, we are not able to compare the severity of the migraine in our population with the studies from France and the UK, because migraine severity was not mentioned in these studies. Triptans were less frequently used in the younger group when compared with the older group (2.4% vs 15.7%). An increase in triptan use by age was also seen in the UK primary care study.
Nausea was present in 71.7%, and vomiting in 49.3% of the participants during a migraine attack. However, antiemetics were only prescribed in a minority of cases by their GP. Although domperidone is an on-label drug for ordinary nausea and vomiting in children, the use of domperidone is not recommended in the DCGP guideline for nausea and vomiting in children suffering from a migraine attack. Nevertheless, it is still unknown if antiemetics are effective as treatment of nausea and vomiting during an acute migraine attack in childhood.
About 12% of the children had used prophylactic treatment before referral. Prophylactic treatment was mainly used in patients from the older group. Unfortunately, there are only a few well-designed trials evaluating prophylactic treatment for migraine in children. The high placebo response in these trials impedes to interpret the efficacy of the drug being tested. It has been reported that the efficacy of topiramate or sodium valproate as prophylactic migraine treatment in children from 5 till 18 years of age is more or less the same.17-19 Another meta-analysis demonstrated only suggestive effectiveness for trazodone and topiramate as prophylactic treatment in children with migraine. A study from the UK reported prophylactic drug treatment by the GPs in 21.4% of the patients between 5 to 17 years of age, with an increase in prophylactic drug use when children were older, confirming the findings of this study. The more frequent prescription of prophylactic treatment in the UK compared with the Netherlands is likely due to differences in GP guidelines. Moreover, differences might be explained by differences in severity of migraine in the 2 studies. However, this could not be retrieved from the UK article.
Migraine characteristics according to the ICHD-II criteria were not associated with the use of medication not listed in the guideline. However, other factors were associated with the prescription of not-listed medication by GPs. In the group with children younger than 12 years, the use of medication not listed in the DCGP guideline tended to be older than the children using only medication listed in the DCGP guideline. In the older group of children, the use of medication not listed in the DCGP guideline was significantly associated with a longer history of migraine or longer lasting migraine attacks. This may be explained by an increased tendency in those patients to visit the GP leading to increased prescription of medication not listed in the guideline.
The limitations of this study are addressed here. This study investigated only a group of children with migraine who are referred to a neurologist. In the Netherlands, only 12% of the children with headache are referred to a pediatrician or neurologist, most of them for migraine. This results in a study population containing only a small fraction of the patients with migraine as seen by GPs. The included patients are more likely to suffer from severe migraine headache or a higher frequency of migraine attacks than those who were not referred. Therefore, the studied patients are more likely to use (listed and not listed in the DCGP guideline) medication. This study did not investigate why the GPs prescribe not-listed medication according to the DCGP guideline.
It has to be noted that the study population represents only 1 regional general hospital. Furthermore, it is a retrospective study. The questionnaires were completed after some time. The amount of time between referral and this study might have influenced the perception of the symptoms and severity of migraine. The questionnaires were completed by patients as well as their parents and the perception of the migraine attacks could be different between parents and patients. However, the frequency of reported symptoms is similar to other studies.21-23 This study was performed in the Netherlands and the guidelines for treating migraine vary between countries.
Despite these limitations, this study provides relevant information on the treatment of migraine in children, which is not available from other sources and could serve to improve the treatment of children with migraine.
To summarize, our study demonstrated that medication not listed in the DCGP guideline is prescribed to children with migraine in primary care. About half of the children with migraine used medication not listed in the guideline of the GPs before referral to a hospital for further treatment of their migraine. Especially older children and children with a longer history of migraine attacks or longer duration of the migraine attacks were associated with the use of medication not listed in the guideline. It is important that the DCGP guideline is supporting the GPs in their daily effort to provide an optimal treatment in children with migraine. The current DCGP guideline is limited in its recommendations and this could be the shortfall to why this DCGP guideline is not always used. A modification of the DCGP guideline is required to support the GPs in the treatment of migraine in children. More prospective research on migraine treatment is required in patients younger than 18 years in the primary care to specify the needed modifications.
Pharmacological treatment in almost 41% of the studied children with migraine in the primary health care setting was not supported by the DCGP guideline
Migraine characteristics such as a longer history of migraine and a longer duration of the migraine attacks were associated with not-listed medication use in primary care setting in patients between 12 and 17 years of age
A modification of the DCGP guideline is required to support the GPs in treating migraine in children
We would like to thank Dr. M.A. Edens (Isala Clinics, Zwolle, the Netherlands) for help with the statistical analyses.
Statement of Authorship
(a)Conception and DesignJan S.P. van den Berg; Ilse F. de Coo; Gosse de Jong
(b)Acquisition of DataIlse F. de Coo; Ronald Zielman
(c)Analysis and Interpretation of DataIlse F. de Coo
(a)Drafting the ManuscriptIlse F. de Coo; Jan S.P. van den Berg
(b)Revising It for Intellectual ContentRonald Zielman; Gosse de Jong
(a)Final Approval of the Completed ManuscriptIlse F. de Coo; Gosse de Jong; Ronald Zielman; Jan S.P. van den Berg