The efficacy of internet training for migraine self-management

Migraine - a chronic disease

Migraine is a common, chronic, incapacitating brain disorder, characterized by
attacks of severe and pounding headache, often unilateral, which lasts 4 to 72
hours. The headache is accompanied by nausea or vomiting and/or intolerance of
light and sound, but also of smell, or movement. About 15% of patients suffer
from migraine with aura, which means that the attack is directly preceded by
focal neurological symptoms in the visual field (scotoma), which last about 20
minutes. Migraine affects roughly 12% of the western population (Goadsby, et
al., 2002; Lipton, et al., 2007, CBS, 2008) with a male-female ratio of about
1:3, which is mostly due to the hormonal changes of the menstrual cycle
(Goadsby, et al., 2002; Lipton, et al., 2007). Despite considerable disability
migraine remains largely underreported and undertreated (Lipton et al., 2000;
2007). About 50% of sufferers seek medical help; in a recent German study this
was 42% (Ratke & Neuhauser, 2008).
Attacks are not confined to the disabling headache but include a
prodromal phase of hours up to days, in which premonitory non-headache symptoms
occur such as tiredness, poor concentration, neck stiffness or irritability
(Giffin et al., 2003; Kelman, 2004) and the patient is susceptible to so-called
migraine triggers, such as stress, female hormones, not eating, weather
conditions or sleep disturbance (Kelman, 2007). Migraine has a strong genetic
background (Russel, 2008), which means that the disease must be managed and
cannot be easily cured. The focus in unravelling its pathophysiology involves
cortical spreading depression in patients with aura and the trigeminovascular
system and central sensitization in explaining the pain, but now extends to the
brainstem and the view that a broad, centrally facilitated dysfunction in
sensory processing is at stake, which also explains the accompanying symptoms
and expands to the prodromal stage (Sprenger & Goadsby, 2009). There remains
controversy regarding the driver(s) of migraine pathogenesis, but abnormal
brainstem function seems to push the pathophysiology of the migraine attack
(Dodick, 2007).

Preventive behavioral treatment - acceptance and required outreach

The treatment of migraine is primarily pharmacological (Sprenger & Goadsby,
2009), with a recent focus on attack prevention (Lipton, et al., 2007; Goadsby
& Sprenger 2010). It is in this area of secondary prevention that behavioural
treatment (BT) comes into play. Its clinical utility was proven in numerous
studies showing that BT can reduce attack frequency - significantly superior to
placebo - with 30-55% (Rains, et al., 2005; Campbell et al, 2000). By now the
preventive usefulness of BT has been internationally acknowledged in
neurological treatment guidelines (for The Netherlands see Nederlandse
Vereniging voor Neurologie, 2007), based on effect sizes of 0.55 and 0.54
(Cohen*s d) of relaxation training and cognitive-behavioral training, the two
most prevailing BT elements. BT aims at the reduction of attack frequency (and
medication use), but also at the improvement of internal control over migraine
and the decrease of impairments due to migraine (Holroyd, et al., 2001). Very
promising for outreach as well as cost-effectiveness was the finding that
home-based BT was as efficacious as clinic-based CBT (Haddock, et al., 1997).
In line with this is the rise of self-management programmes in chronic disease
(Bodenheimer, et al., 2002; Farrel, et al., 2004) and evidence that lay
trainers with a chronic condition can deliver this type of intervention
(Foster, et al., 2007). The present focus on self-management in chronic
disease, and on self-efficacy (Farrel, et al., 2004) or empowerment of patients
(Samoocha, et al., 2010), is driven by the societal urgency to solve the
growing gap - induced by an escalation in chronic disease - between care supply
and care demands (Irsel van, 2006). Thus, effective self-management is at stake
too in chronic migraine, as is its outreach and cost-effectiveness.

Self-management for attack prevention and training offered by lay trainers

We therefore developed a self-management training (SMT) rooted in
evidence-based protocols for BT in migraine (McGrath, Holroyd, & Sorbi, 2000;
Sorbi, & Swaen, 2000a/b; 2004). SMT includes patient education, detailed
self-monitoring, a relaxation training protocol and a limited dose of
cognitive-behavioral training. SMT has two main focal points: (1)
identification and modification of prodromal features of the attack
(premonitory symptoms and migraine triggers), and (2) direct employment of
voluntary body relaxation and cognitive-behavioral self-regulation skills to
counteract the dysregulation and prevent attack occurrence in the prodromal
stage (Mérelle, 2008a-c). Both steps are not easily achieved given the habitual
responses of many migraine patients. One is to focus on the hell of the attack
while ignoring the prodromal features that are comparatively inconspicuous; the
other is to increase effort and exertion when facing an imminent headache,
which promises to get things done before the pain strikes but in fact is vastly
counterproductive (Sorbi, 2010).
If brainstem dysfunction drives attack occurrence, then it is plausible
that stabilization of arousal and buffering the susceptibility to sensory and
other stimulation makes sense and potentially supports attack prevention. This
is what SMT intends through the early detection of warning signals and direct
employment of behavioural self-regulation. This self regulation is crucial when
attacks are close, but also seems relevant in between attacks, given that
evoked and event-related potential studies in migraine revealed an interictal
lack of habituation of the sensory cortices (Schoenen, et al., 2006). Last,
these skills help to regulate the proposed sympathetic hyper-function in
migraine (Ludwig, et al., 2006).
SMT was extended from the clinic to the open population, and from
education to training of prevention skills, while involving trained lay
trainers as positive role models to provide SMT at home to small groups of new
patients with chronic migraine (Mérelle, 2008). It was offered to who suffered
from one to six attacks but less than 15 migraine days per month (Bigal, et
al., 2008), and did not present with medication overuse, evident
psychopathology or complex comorbidity (Mérelle, 2008a-c).

Outreach, efficacy and acceptance of migraine self-management training

SMT was tested in a large randomized controlled trial with 13 patient trainers
and 30 groups of new patients with an extended follow-up of six months (N=95).
Participants were recruited through the patient organisation (11%), multimedia
in the open population (87%) and headache specialists (1%). The outreach was
good given the large response (N=607). Of 264 patients who provided informed
consent 48% could be included on the basis of diagnostic questionnaires and a
four-week headache diary. After inclusion the drop-out rate was very
reasonable: 15% of participants did not complete the training, 10% did not
provide the follow-up measurements six months after SMT (Mérelle, 2008b,c).
The efficacy of the present SMT application in reducing attack frequency was
modest (-23%, effect size: 0.6)
compared to clinic-based BT, but was comparable to that of training by lay
trainers for other chronic diseases (Foster, et al., 2007). Migraine-specific
quality of life significantly increased over time and SMT strongly improved
perceived control and self-confidence in attack prevention that remained stable
during six months follow-up (Mérelle, 2008a-c). The acceptance of SMT and of
the patient trainers who provided it was very good (Mérelle, 2008a-c).
Therefore the Dutch Society of Headache Patients urged to extend the provision
of SMT by offering it through the internet, which would also enhance SMT
accessibility by making it independent of time and place or travel distance.
This plea of the patient organisation was reinforced by our introduction during
the study of an online method to support SMT directly and in real life
(Mérelle, 2008; Kleiboer et al., 2009).

Use of mobile SMT support

Online Digital Assistance (ODA) employs mobile monitoring and coaching with
smartphones and wireless internet. Mobile monitoring consists of random
prompting the keeping of an electronic diary to record migraine, prodromal
features and self-management. This method is scientifically sound and generates
valid real time assessments unbiased by retrospection (Bolger, et al., 2003).
Mobile coaching consists of direct personal feedback in three parts on the
smartphone screen concerning (1) actual state of prodromal features and attack
risk displayed by a traffic light with brief arguments from the present diary,
(2) advice and tips for preventive action, and (3) positive reinforcement of
self-management supported by an emoticon.
ODA runs on advanced and protected clinical software and was employed
in 44 patients to reinforce self-management during the last part of SMT. ODA
was feasible and its acceptance when offered in training was high. ODA was
found useful and supportive while the burden was low, and according to the
participants sustained attack prevention (Sorbi, et al., 2007; Kleiboer, et
al., 2009). However, SMT with ODA did not yield better improvements, compared
to SMT only (Kleiboer, et al., 2009). In this respect it was considered that
surplus effects up and above the effects of training that served the same
purpose require larger groups for statistical power. In addition, two
alternatives were taken into account to render ODA more profitable than when
applied during training. First, since SMT is intensive and its effect
consolidates over time, ODA could be employed several months after completion
of SMT to reinforce the maintenance of self-management. Second ODA may be
prolific as a method in its own right to promote self-management or as a means
to prepare subjects for SMT (Mérelle, 2008; Kleiboer, et al., 2009).

Preparation of the present study: development of SMT through the internet with
independent options for mobile monitoring and coaching

Development of an internet SMT for migraine was in perfect alignment with the
current general trend in internet intervention, which evolves rapidly and
successfully (Barak, et al., 2008; Cuijpers et al., 2008, Webb, et al., 2010).
Supported by a national grant we developed the technology (content management
system, CMS) for the advanced and protected delivery of screening, training and
effect measurement through the internet, and we translated the materials to
suit the internet medium in eight SMT lessons. The internet training
*MyMigraine* was completed in 2009 after its feasibility and acceptance was
successfully tested in 10 new patients and evaluated by 6 patient trainers who
participated as SMT experts (Sorbi, 2009; Sorbi & van der Vaart, 2010). The
patient ratings consistently confirmed the clarity, instructiveness, importance
and easy execution of all lessons, and after training the patients were
positive about user-friendliness and clarity, training content and benefits,
and their general satisfaction with the internet-provided SMT. The expert
patients considered MyMigraine instructive, captivating and fun to work with
and were highly positive regarding the web application, digital support and web
adaptation of the protocol (Sorbi & van der Vaart, 2010). One point of
consideration is growing evidence that minimal support, contact and coaching
during internet intervention is essential to induce benefit and prevent
substantial drop-out (Andersson, 2009; Donker et al., 2009). Therefore
MyMigraine will be supported by weekly e-mail contact according to current
empirical standard (Wammerdam, et al., 2008; Carlbring et al., 2007), which
takes 20 minutes per participant per week.
After MyMigraine was finalized the ODA application was integrated into
the CMS and evaluated regarding feasibility, compliance and acceptance in five
migraine patients (Sorbi, et al., 2010). ODA as part of the CMS can be employed
on call, that is, when the researcher decides to do so either before, during or
after completion of the SMT.

The present study: is SMT through the internet efficacious in chronic migraine
and does mobile monitoring and coaching contribute to maintain benefits or to
prepare for training?

The primary focus of the present study is on the efficacy - post training and
after six months follow-up - of SMT for chronic migraine when SMT is provided
through the internet. This is a challenge because, if established, SMT could be
made widely available in Dutch health care. In addition, it is conceivable that
the present SMT with its surrounding technology could also serve as an
interface between the patient organisation and the Society of Dutch Headache
Centres and function as a preventive aid in the public health arena, involving
professional input (and costs) on indication only. This would be in line with
the present web-driven shift in health care, now coined as *health 2.0*, which
substantially extends the impact of patients and their control over care (RVZ,
2010). And last, next studies could focus on translating the SMT and making it
available to other (European) countries.
Other aims of the present study concern three issues: the evaluation of
change induced by mobile monitoring and coaching employed either after or to
prepare for SMT; the comparison of SMT provided through the internet with
former home-based SMT offered by patient trainers, and first steps in the
economic evaluation of internet-delivered SMT compared to a waitlist for adults
with chronic migraine (Warmerdam, et al, accepted; Edwards, et al., 2010).

Primary Objectives

The primary objectives of this study are to establish:
1. The efficacy of internet-delivered SMT for chronic migraine - compared to
waitlist control- regarding the post-training improvement of migraine attacks
and headache days per month, as well as two measures reflecting empowerment,
e.g. self-efficacy (the confidence that one can take action to prevent attacks)
and internal control (the belief that migraine occurrence and relief are within
one*s own power);
2. The maintenance of internet-delivered SMT benefits six months after training
regarding all measures outlined above;
3. The utility of mobile monitoring and coaching in (a) supporting the
maintenance of internet-delivered SMT benefits and in (b) preparing for this
type of training.

Secondary Objectives
4. Secondary measures of efficacy are the intensity and duration of the
attacks, migraine-specific disability and quality of life;
5. The economic evaluation of internet-delivered SMT compared to waitlist.

We conduct a parallel-group randomised controlled trial with a delayed
intervention design, conform the study that established the efficacy of
home-based SMT delivered by patient trainers (Mérelle, 2008a-c). Measures are
taken at baseline (T1: pre-test), after training (T2: post-test) and six months
following the post-test (T3: 6-mths follow-up). Inclusion criteria are conform
the former study and established with an headache diagnostic indicator
(ID-Migraine), four weeks of headache monitoring according to the criteria of
the International Headache Society (IHS, 2004) and medical consultation in case
of doubt, and an extended questionnaire to establish psychopathology and
psychological comorbidity (SCL-90R). A training group (TG) and waitlist control
group (WL) will enter the study in three cohorts.
The decision to use a waitlist control group - rather than a psychological
placebo condition - is based on the difficulties previous studies had in
realizing a sound placebo procedure (Rains, & Penzien, 2005; Blanchard et al.,
1990). In contrast to the former study, however, we will extend the control
period to T3 to cover the 6-months follow up.
The assessments of the primary and secondary measures are in line with
the study of Mérelle (2008). A headache diary according to the IHS (2004) is
employed for four weeks at T1, T2 and T3 (and is kept during training as well),
which allows to assess the primary outcome measure migraine attacks, and which
yields accounts of headache days as well as the use of abortive, analgesic and
prophylactic medications for headache. In the former study a paper diary was
employed, which is presently replaced by an electronic headache diary measuring
the same variables. Its feasibility and acceptance was successfully established
in 29 migraine patients who kept it for 80 days at average (van Silfhout et
al., 2010). A large advantage is that migraine attacks according to the IHS
(2004) are identified by the software and are graphically represented over time
in the so-called migraine monitor. This improves the view on progress for
participants and investigators; it also excludes potential errors due to the
calculation by hand required with the former paper diary.
The Dutch versions of four international and migraine-specific questionnaires
are employed at T1, T2 and T3 to assess the primary measures for empowerment
(self-efficacy and perceived control) and the secondary measures for migraine
disability, quality of life, (see section 5.2 in the research protocol). In
addition, participants fill in SMT evaluative questionnaires at T2 and T3, as
well as an ODA evaluative questionnaire after conclusion of the three weeks of
online monitoring and coaching, which are derived from the former study.
Mobile monitoring and coaching will be offered halfway in the 6-months
follow-up period to 50% of the participants in both groups (ODA+; N=60). In
this way a 4x4 delivery matrix is realized in which ODA+ and ODA-, as well as
ODA pre-training (WL) and ODA post-training (TG), is systematically varied with
30 participants per sub-condition. This enables the analysis of whether ODA
supports the preparation for - and/or the effect maintenance of - internet
delivered SMT (3rd objective of the study).
The data acquisition will be completed after 30 months and publications
will be finalized in the six months that follow.

The internet training *MyMigraine* is grafted on an evidence-based SMT
protocol, tailored to the format of an Internet-delivered intervention.
Potential participants can enrol via the website www.mymigraine.nl.
Subsequently they are asked to fill in the Symptom Checklist 90R (SCL-90R), the
ID-migraine (diagnostic scale), and a 4-week headache diary. When decided that
the training is suited for the participants they receive authorization and can
commence with the training. The training consists of BT strategies that are
central to attack prevention in migraine: 1) identification and modification of
triggers and affective, cognitive and behavioural premonitory symptoms; 2) use
of physiological and cognitive-behavioural self-regulation skills. The main BT
techniques for migraine are relaxation training (mean ES = 5.5) and
cognitive-behavioural training (mean ES = 5.4) (Nederlandse Vereniging van
Neurologen, 2007) and the SMT includes a complete autogenic relaxation
training, supplemented with cognitive behavioural strategies such as goal
setting, self-reinforcement and positive thinking.
The online SMT consists of 8 lessons that are spread over 8-12 weeks.
The lessons will take the participant approximately 1 hour to complete and
preferably take place once every 7 to 10 days. The average time investment is
approximately 30 minutes per day for two daily relaxation exercises (30 minutes
in the first half and 10 minutes in the latter part of training) and
cognitive-behavioural homework. Headache self-monitoring requires a few minutes
per entry.
The first lesson acquaints the participants with the principles of the
training and provides health education on the subject of behavioural attack
prevention. Lessons 2-4 focus on 1) detection of premonitory features of the
attack (migraine triggers and premonitory warning symptoms) by daily
monitoring, and 2) acquisition of relaxation skills by stepped practice of
autogenic and breathing exercises twice per day at home. Sessions 5-8 focus on
1) application of relaxation skills and other proactive strategies under the
condition that premonitory symptoms or triggers prevail, while continuing the
daily exercises at home, and 2) the formulation of personalized prescriptions
for migraine and health with a focus on individualized target conditions of
attack risks and specific actions or lifestyle changes to prevent attack
occurrence. Evaluation takes place in lesson 8, which includes the
specification of individual goals and actions to maintain the self-management
skills.
Participants in the waiting-list condition (WL) continue with their
current treatment, i.e. care as usual, while keeping the electronic headache
diary. The WL-group receives the training after the waitlist period.
To keep participants informed they will receive an annual (online)
newsletter concerning the research project.
After completion of the training in the TG condition and before the
start of training in de WL condition, half of the participants are offered 3
weeks of mobile monitoring and coaching with a smartphone that connects through
wireless internet to the content management system of the internet training.
The smartphone contains a digital diary, of which selected ratings are the
target of direct and personalized online feedback provided by a staff member of
the project or by trained master students under supervision. The diary keeping
takes several minutes to fill in questions about 1) migraine and medication
use; 2) prodromal features of attacks; 3) voluntary relaxation and other
self-regulation strategies for attack prevention. Four beeps are randomly
programmed within 2.5 hour blocks separated by 0.5 hour each day to signal that
the diary should be filled in. In addition to the beep diary participants fill
in shorter diaries upon waking up and before going to sleep.
Mobile coaching is provided by direct feedback to the digital diary
within a template on the smartphone screen (see figure 1in the research
protocol) that consists of three parts: 1) current attack risk with markers
from the diary, accompanied by a traffic light, 2) tips and advice for
self-management derived from the training and 3) positive reinforcement with
brief encouragements underscored by an emoticon (smiley). In case of moderate
or severe headache the device automatically provides the option to stop the
diary.

The benefits of self-management training for migraine, and of mobile monitoring
and coaching to support it, have been outlined (see the section on the
background of the study). There are no foreseeable risks for the participants
in this study. This was confirmed by the absence of adverse events in the
former study of home-based SMT provided by patient trainers (Mérelle, 2008;
reviewed by the METC of the Erasmus MC in Rotterdam, code P03.1227L). The time
participants have to invest in the training program was and is substantial ,
but this did not translate into a large percentage of drop-out. Participation
is voluntary and participants are fully informed from the start about the
nature of the training. Strong efforts were made to make the internet
application user-friendly and captivating, which was confirmed by six expert
patients involved in the feasibility pilot study (Sorbi & van der Vaart, 2010).
The burden of travel time and expense is relieved by offering the intervention
through the Internet, which has the additional advantage that it can be
followed at a time and place of the participants own choosing.