This study is a single-blinded, multicentre, pragmatic clinical trial, with two parallel groups assessing the potential superiority of a multimodal MT treatment over usual care by the GP. We will include a four weeks run-in period to provide accurate migraine frequency data prior to enrolment. The treatment will last 12 weeks with follow-up measurements at 12, 26 and 52 weeks (Figure 1). The study adheres to the guidelines of the International Headache Society (IHS) for controlled trials in patients with migraine regarding inclusion criteria, outcome measurements and statistical analysis [9].
Parallel to the RCT we will conduct a prospective cohort study with migraine patients with a strong preference for MT treatment who do not want to be randomized. The aim of this parallel group is to explore the differences in patient characteristics at baseline between the randomized trial and the cohort study. Patient’s expectations with regard to recovery will be assessed in both studies and the relationship between expectations and effect of the treatment, i.e. the primary outcome measures, will be analysed. This information is necessary to better understand the generalizability of the study results. The participants in the cohort study will be treated with MT; treatment and measurements will be identical to the treatment procedure and measurements used in the RCT. All measurements will take place by a trained research assistant at one location. Participants will be given usual care by their own GP.
The design and protocol of the study have been approved by the medical ethics committee of Amsterdam University Medical Centres (location VUmc) and registered in the Dutch Trial Register (NL7504) The SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) for this research is available in Additional file 1 [28]. Information of participants will be handled according to EU General Data Protection Regulations, and according to the guideline of the Central Committee on Research Involving Human Subjects (CCMO) in the Netherlands, to protect participant confidentiality. The study is monitored by the Clinical Research Bureau (CRB) of Amsterdam UMC location VUmc.
Population
Participants will be recruited by the participating general practitioners (GPs) working in an urban area of Hoofddorp, The Netherlands. During consultation, the GP will provide oral and written information about the study and will invite the patient to participate. If the patient is interested in participating in the study and consents in providing contact details to the researcher, the GP notifies the researcher by email. The researcher will provide additional information about the study to the participant, followed by a telephone interview after one week to answer possible questions and check inclusion and exclusion criteria.
Information of the study will be provided by posters and folders in participating general practices and at the website of the coordinating healthcare centre.
Inclusion criteria
Eligible participants are between 18-65 years of age and should have had migraine attacks for more than one year, according to the diagnostic criteria of the International Classification of Headache Disorders (ICHD) III [29]. A GP or neurologist should have established the diagnosis migraine, and the frequency of attacks should be two times a month or more. Co-occurrence of tension-type headache is allowed if the participant can clearly distinguish this headache from migraine. Participants will only be included if they have concomitant neck pain between migraine attacks or during an attack. The use of prophylactic medication is allowed if migraine is stable and medication use has not changed in the last three months. Furthermore, participants have to be able to read and write Dutch.
Exclusion criteria are (suspected) malignancy, pregnancy, cerebrovascular disease, degenerative central nervous system diseases, medication-overuse headache, a current diagnosis of depression or other severe psychiatric disease, rheumatoid arthritis, serious or systemic infection, fever, or change in medication for migraine within three months before the study, and having received MT treatment for migraine up to three months prior to the start of the study.
Data collection
Participants are asked to keep a headache diary and will be instructed on the way how to report, in order to obtain baseline data for migraine characteristics, and will receive an appointment with the research assistant after four weeks. Headache diaries will be provided on paper. On each day, participants are able to report: no headache, tension-type headache or migraine, medication use and absence of work because of migraine. Before baseline measurement, the research assistant will check inclusion and exclusion criteria and ask for written informed consent. Four weeks before the follow-up appointment, participants will be asked by email or telephone to start filling out a headache diary.
At all measurements, the data will be collected electronically (Castor EDC). The secured electronic program includes data validation checks and a data audit trail according to good clinical practice standards. Only the researchers and the research assistant will have access to the data. Information of participants will be handled according to privacy regulations of the Amsterdam Public Health Quality Handbook (http://www.emgo.nl/kc/privacy/).
Baseline assessment
Baseline assessment will include the registration of demographic variables (age, gender, education and profession), migraine characteristics according to ICHD III criteria, other physical complaints and chronic diseases.
Expectations regarding the effectiveness of treatment will be measured on a 7-point rating scale (range from no result (0), to excellent result expected (6)). Patient's preference for treatment will be administrated (preference for usual care, MT or no preference). Other outcome measures include disability (Headache Impact Test questionnaire (HIT-6)) [30], allodynia (allodynia questionnaire) [31], pressure pain thresholds [32] and neck flexor muscle endurance [33] (see below for details). Figure 2. shows all outcome measures and assessments.
Randomization
After baseline measurement, randomization will take place with a 1:1 allocation ratio. An independent statistician who has no involvement with the clinical investigators will generate a random sequence of numbers before the start of the study. The research assistant who is blinded for the randomization sequence will supply sealed and numbered envelopes. In the presence of another administrative assistant, the participant will open the sealed envelope, and an appointment will be made for treatment by either the participating GP or a manual therapist. Participants will be invited to a parallel cohort study if a strong preference for MT treatment keeps them from agreeing with randomization.
Blinding
Allocation of participants is concealed from the researcher and the research assistant who performs all measurements. An independent statistician will carry out the statistical analysis and review the interpretation of the results. For obvious reasons, participating patients, GPs and manual therapists cannot be blinded to treatment.
Usual care
Participants assigned to the usual care group will be treated by their GP. The GP will treat participants as usual, based on the recommendations of the practice guideline for headache of the Dutch College of General Practitioners [34]. The general practitioner (GP) provides lifestyle advice and, if necessary, prescribes medication. The recommended treatment consists of acute medication for a single attack or prophylactic medication when the migraine attacks occur two times a month or more [5]. The GP will evaluate the treatment in consecutive appointments. Participating GPs will be informed about the research protocol by the researcher during a one-hour meeting.
Intervention
The multimodal manual therapy treatment aims at restoring cervical function in order to reduce nociceptive cervical afferent output. The treatment will include manual pressure techniques on the trapezius muscle and upper cervical/suboccipital musculature to decrease neck pain intensity and cervical muscle tenderness [35]. Neck muscle strength will be trained, by giving low load craniocervical muscle exercises and correcting sitting and standing posture [36]. The selected spinal mobilizations are low and high-velocity techniques of the cervical and thoracic spine. To create a protocol that will be feasible for all participants, no high-velocity thrust manipulations of the upper cervical region (C0-3) will be applied in the study, due to possible individual risk factors associated with serious adverse events [37] [38].
Experienced manual therapists will be trained in the treatment protocol prior to the study. The treatment protocol provides recommendations of techniques that can be used; the treating manual therapist decides which techniques will be included, depending on the condition of the participant. The applied techniques will be documented for each session on an evaluation form. Instructions on posture and home exercises will be provided to the participants in booklets. The MT intervention will consist of a maximum of 9 sessions of 30 minutes starting with treatment once a week, followed by once every other week during 12 weeks.
During the 12 weeks of treatment, participants will be asked not to make use of additional therapies or medication for their migraine. At all follow-up measurements possible use of additional therapies and medication will be asked for and registered.
Primary outcome measure
The primary outcome of the study is the number of migraine days, recorded by the participant in a headache diary during all follow-up assessments. [9]. A migraine day is defined as a day with migraine characteristics according to the IHS classification ICDH III for longer than four hours, or a headache that resolves with the intake of triptans or ergotamine within two hours of intake [29].
Secondary outcome measures
The secondary outcome measures are:
- Number of migraine attacks per four weeks, recorded in a headache diary during the four weeks before follow-up measurements [9]. Migraine attacks will be considered as separate attacks if 48 hours without headache is reported in the headache diary
- Intensity of migraine, assessed on an 11 point numerical rating scale (0 = no pain, 10 = most severe pain) [39].
- Intensity of neck pain, assessed on an 11 point numerical rating scale (0 = no pain, 10 = most severe pain) [39].
- Medication use in number of doses per 4 weeks of simple analgesics (g., paracetamol), NSAIDs, acute migraine medication (triptans and ergotamines) or prophylactic medication. Participants are asked to report changes of medication to the research assistant at all follow-up measurements.
- Responder rate will be measured by the number of migraine days before vs. after treatment, dichotomized into ≥50% reduction or not [9].
- Disability, assessed by the HIT-6 questionnaire. The HIT-6 consists of 6 questions measuring pain intensity, social functioning, role functioning, vitality, cognitive functioning and psychological distress on a 5-point ordinal rating scale (never to always). Internal consistency is considered high (Cronbach's alpha 0.82 to 0.90), and test-retest reliability is fair (ICC 0.77) [30]. The Dutch version of the HIT-6 questionnaire has shown to be a valid and reliable tool to measure the impact of migraine [40].
- The endurance of the neck flexor muscles will be scored as the number of seconds the participant can raise his head from the table when lying in supine position as described by Harris et (2005). Harris et al. reported good to excellent intra-tester reliability (ICC 0.82–0.91) and moderate inter-tester reliability (ICC 0.67–0.78) [33].
- Cutaneous allodynia (CA) will be evaluated with the 12-item allodynia symptom checklist. This questionnaire consists of 12 questions about cutaneous hypersensitivity in the cervical cephalic The participant can score yes, no, or not applicable. Allodynia symptoms and score on CA severity are defined in the following categories: none (0-2), mild (3-5), moderate (6-8) and severe (9 or higher) [31].
- We will perform algometry, by measuring pressure pain thresholds (PPT) with a Wagner FDK algometer at the upper trapezius muscle (at the midpoint between C7 spinosus and the acromion), the suboccipital muscles and the anterior tibial muscle. The PPT measurement will be repeated three times at each point, and a mean score will be calculated. Algometry has demonstrated excellent intra-tester reliability (upper trapezius test-retest ICC 0.83, 95% CI 0.69–0.91), and excellent inter-tester reliability (upper trapezius ICC 0.89, 95% CI 83–0.93) [32].
- Participants will be asked to report global perceived effect on a 7 point rating scale (0 = much worse to 6 = much better). Disability due to attacks will be assessed on a 5 point rating scale (0 = no disability and no medication to 4 = fully disabled even with medication). Also, use of healthcare resources and absence of work will be reported.
- All adverse events will be administrated for both treatments at all follow-up measurements.
Statistical analysis
Baseline characteristics will be presented in percentages for categorical variables, and in means and standard deviations for continuous data, using descriptive statistics. The distribution of the data will be evaluated using histograms and QQ-plots. The outcomes will be adjusted for baseline differences. The outcomes of the total follow-up period, including baseline data, will be examined with a linear mixed model analysis. Differences between groups will be reported and shown in tables. Differences between the cohort group and the RCT groups (separate and combined) will be analyzed with Student T-tests (continuous data) and Chi-squared tests (nominal data). For non-parametric data, the Mann-Whitney U test will be used. The primary analysis will be by intention-to-treat. Additionally, a per-protocol analysis will be carried out to assess the effect in participants who adhered to the protocol. Protocol adherence will be defined as staying in the allocated treatment group during the 12-week treatment period; for MT treatment, participants have to complete at least six sessions. In the ‘usual care’ group, participants who receive MT during the 12-week period will be excluded from the per-protocol analysis. Effect sizes will be computed for normally distributed outcomes. Statistical analysis will be carried out using SPSS version 23 (IBM Corporation, Armonk, NY).
Sample size
Taking pilot study results as the basis for our calculation, we assume an average frequency of migraine days in the 4 weeks before the 26th-week measurement point of 4.2 days (SD 2.4). As we want to detect a difference in the reduction of the number of migraine days of at least 25% between groups, with a two-sided significance level of 0.05 and a power of 0.80, each group will have to include 83 evaluable persons. Taking into account a 15% loss to follow-up, a total of (100/85)*83*2=196 participants will have to be enrolled into the study, 98 per group.
To ensure the enrolment of the required number of participants over an estimated period of two years, we will recruit 44 GPs and four manual therapists to participate in the full trial.
Feasibility of the study
We performed a pilot study to assess the feasibility of the measurements, the treatment protocol and randomization procedures. The pilot study concerned 24 possible participants in eight weeks (October 2015 to December 2015); 11 participants fulfilled the inclusion criteria.
Two out of 13 excluded participants had a strong preference for manual therapy treatment and, therefore, were excluded from randomization. Other reasons for exclusion were: no migraine according to the IHS criteria, low frequency of migraine and participants with GPs who did not participate in the pilot study. The research protocol was evaluated by questionnaires and in personal meetings with the participating manual therapists, GPs, research assistant and participants. GPs and MT reported no problems with adhering to the protocol for measurements and treatment. The results of the pilot study showed that the treatment protocol and procedures were feasible and that the participants tolerated both treatments well.