Non-pharmacological educational and self-management interventions for people with chronic headache: the CHESS research programme including a RCT

Article history

The research reported in this issue of the journal was funded by PGfAR as project number RP-PG-1212-20018. The contractual start date was in January 2015. The final report began editorial review in April 2021 and was accepted for publication in February 2022. As the funder, the PGfAR programme agreed the research questions and study designs in advance with the investigators. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The PGfAR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Copyright statement

Copyright © 2023 Underwood et al. This work was produced by Underwood et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaption in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

2023 Underwood et al.

Background

Headaches are second only to low back pain as a global cause of years lived with disability, accounting for 6.4% of the total years lived with disability. 1 Only dental caries are more common than migraine and tension-type headaches. 1 Headaches are the most common neurological disorder treated in primary care, accounting for around 3% of general practitioner (GP) consultations; however, 70% of these patients do not get a formal diagnosis. 2,3 Two-thirds (64%) of people seen in a specialist headache clinic have not had a headache diagnosis from their GP. Of those with migraine, fewer than half have been offered specific migraine treatment. 4

For many people with headache disorders their symptoms are intermittent and they can be managed with symptomatic treatment as required. However, for a substantial group of people, headaches become a chronic disabling disorder. This group contributes disproportionally to the health burden, and economic cost, of headache disorders.

Treatment guidelines for headache disorders are typically formulated in a biomedical framework, with the main focus being on drug treatments for those with a diagnosed headache disorder. For example, the only non-pharmacological intervention recommended for people with chronic tension-type headache or chronic migraine in the 2012 National Institute for Health and Care Excellence (NICE) guidelines5 published was acupuncture. There is a need for more non-pharmacological treatments to help those living with headache disorders.

In 2015, when we started this programme of work, very little was known about how best to support those who have developed a chronic problem consequential to a primary headache disorder. We anticipated that any supportive self-management programme would need to include helping more people to get a headache diagnosis, avoiding medication overuse headaches, providing medication appropriate for the headache type and supportive self-management for a chronic painful disorder.

Headache researchers usually expect those individuals studied to have an established headache diagnosis prior to study entry. 6,7 This work, however, started from the premise that most people with frequent headaches do not have an established diagnosis, and was developed from the perspective of headaches as a chronic disorder. There is not, however, a recognised clinical entity of chronic headaches. In epidemiological studies researchers use a definition of chronic headaches based on the definition of chronic migraine or chronic tension-type headache, specifically headaches on ≥15 days per month for the previous 3 months. 8–11 This matches neither the conventional definition for chronicity of pain used by International Association for the Study of Pain (IASP), of pain persisting for greater than 3 months,12 nor the definitions of chronicity used for other headache disorders, for example chronic cluster headache. 9 In addition, the chronicity of headaches defined in this manner is labile. 9,12 In a community study of migraine,13 three-quarters [386/526 (73%)] of people with chronic migraine had one or more 3-month period in the following year when their headache frequency was consistent with episodic (<15 headache days per month) rather than chronic migraine. 13 Nevertheless, this epidemiological definition is a useful shorthand to describe our population of interest: people living with headaches affecting them on most days. Our overarching aim was to develop and test a supportive education and self-management group intervention implementable in primary care for people with chronic headaches. This is an area with little previous methodological work and an absence of substantial previous UK experience of recruiting people from primary care for studies of headache interventions. A broad programme of research was needed to set the scene for our randomised controlled trial (RCT).

Key areas of uncertainty were as follows:

1. Was it possible to recruit people from primary care who met our definition of chronic headache?

2. How easy would it be to identify people who might have other headache types not suitable for our intervention package?

3. What is the patient experience of living with chronic headaches (including chronic migraine)?

4. What should be the content of the self-management support intervention?

5. What format should be used for the delivery of the self-management support intervention?

6. How acceptable would our intervention package be to people living with chronic headaches?

7. How should we measure outcomes?

These were all addressed in the phase 1 feasibility study. In phase 2 we ran a full RCT, with a cost-effectiveness analysis and an embedded process evaluation.

Phase 1: trial feasibility

The trial feasibility phase was made up of four core areas of work, each of which mapped onto our work packages (WPs):

1. WP1 – developing a strategy to identify people with chronic headaches from primary care.

2. WP2 – developing and evaluating a telephone headache classification interview.

3. WP3 – developing and evaluating an education and self-management support intervention for people living with chronic headaches.

4. WP4 – selecting the most appropriate patient-reported outcome measure (PROM).

The trial feasibility phase aimed to answer the questions of what can be done, what should be done, and how best it can be done for a future RCT. Here we describe these packages of work and how these informed the trial feasibility (see Figure 1).

FIGURE 1.

Components of the trial feasibility phase. PPI, patient and public involvement. Reproduced with permission from White et al. 14 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The figure includes minor additions and formatting changes to the original.

Copyright © 2023 Underwood et al.

Recruiting feasibility sample (work package 1)

Objective: to develop and test strategies for recruiting people living with chronic headaches from primary care.

To determine recruitment feasibility, we needed to:

1. Develop a strategy to identify people with chronic headaches from primary care. Our scoping work identified that the standard clinical terminology system (Read codes) for coding chronic headache disorders in general practice was rarely used. We developed a search strategy incorporating age, consultation for headaches and prescription of headache-specific medication to identify our target population.

2. For any future trial, we needed to be sure we could recruit practices. To test this, during the feasibility phase we aimed to recruit practices in the West Midlands. With the help of our local Clinical Research Network (CRN) we recruited 14 practices, giving us a total practice population of 128,634. A detailed account of the recruitment process is given in our published paper (see Appendix 1). 14

From the practice population of 128,634 we obtained informed consent from 131 participants (see Figure 2). 14 Participants’ mean age was 49 years [interquartile range (IQR) 38.5–58 years, standard deviation (SD) 13.3 years], 108 (82%) were female and 125 (95%) were white. These participants consented to:

FIGURE 2.

Practice and participant recruitment for the feasibility study. Reproduced with permission from White et al. 14 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The figure includes minor additions and formatting changes to the original.

Copyright © 2023 Underwood et al.

1. completing an electronic (smartphone/web-based) headache diary (a paper version was also available) for 3 months

2. filling out a baseline, 2-week and 3-month questionnaire

3. taking part in two telephone interviews

4. the research team sharing the headache assessments with the participants’ GPs.

They also agreed to potentially being invited to two further studies:

1. an interview study to explore the experience of living with frequent headaches, how we might refine our proposed support programme, and what outcomes are important to people living with frequent headaches

2. a pilot study in which we invite people to take part in our chronic headache education and self-management support programme.

The sample size for the feasibility study was driven by the need for sufficient data to measure the level of agreement of the classification interviews in WP2. The original sample size was 170; this was revised to 153 paired interviews owing to a change in planned analyses approved by the Programme Steering Committee (PSC) and the funder in March 2016. Later, the PSC agreed that recruitment could stop at 131 participants after reviewing the agreement analysis for the first 100 paired classification interviews. This cohort of 131 participants provided us with a sampling frame for the feasibility work; full details are in our published paper (see Appendix 1). 14

Developing and evaluating a structured telephone interview for diagnosing common chronic headache disorders (work package 2)

Objective: to develop and evaluate a brief diagnostic interview to support diagnosis for people with chronic headaches, focusing on the diagnosis of the three common chronic headache disorders – migraine, chronic tension-type headache and medication overuse headache.

Within this WP it was important not only to be able to identify the population of interest for the main trial, but also to feel confident that those who have other headache types not suitable for our trial were appropriately identified, and referred for relevant support. We first reviewed the literature on diagnostic tools.

Consensus conference

The aim of the day was to reflect on the evidence from the review and draw on the expertise of the delegates to help inform the content of the Chronic Headache Education and Self-management Study (CHESS) nurse telephone classification interview.

The purpose of the meeting was to develop a classification interview that would:

1. confirm study eligibility – participants aged ≥18 years with chronic headache, defined as a headache on ≥15 days per month for at least 3 months

2. classify the participant’s headache type as part of the active intervention to inform treatment and advice (this was done during the face-to-face part of the intervention).

Twenty-six delegates attended (10 neurologists with specialist interest in headaches, three general neurologists, five headache specialist nurses, one GP with a special interest in headaches and seven people living with chronic headaches) attended our ‘Classification Consensus Conference’ at the University of Warwick in August 2015.

They were split into four multidisciplinary groups. Using facilitated discussions each group was asked to address the following four questions:

1. What do we need to know from a person to exclude secondary headaches?

2. What do we need to know from a person to exclude primary headaches other than chronic migraine and tension-type headache?

3. What do we need to know from a person to distinguish between chronic tension-type headache and chronic migraine?

4. What do we need to know from a person to identify medication overuse headache?

The facilitators aimed to get consensus on discussed items; when there was uncertainty, items were taken to a plenary session in which they were further discussed followed by voting to gain consensus. Further work by the research team developed and refined the final classification logic model underpinning the classification interview (see Figure 3).

FIGURE 3.

Classification logic model. a, Tension-type headache, has mild nausea. NSAID, non-steroidal anti-inflammatory drug. Reproduced with permission from Potter et al. 16 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The figure includes minor additions and formatting changes to the original.

Copyright © 2023 Underwood et al.

The classification logic model is presented as a linear process. However, it was developed to be completed non-sequentially as information was obtained during the interview and it is not a diagnostic algorithm.

Having developed the classification tool, the next phase was to validate the tool by training nurses to use the tool.

Developing and evaluating the intervention (work package 3)

Objective: to develop and pilot an education and self-management support intervention for the management of common chronic headache disorders (the CHESS intervention) that is both theoretically informed and based on best evidence.

Details of our intervention development are in our published paper (see Appendix 4). 19

We first reviewed the existing literature to understand the experience of chronic headaches from the patient perspective, what content and approaches might be effective for this population and what modifiable prognostic factors exist to be targeted in future interventions.

Testing and refining the intervention package

Having developed the intervention package and a facilitator manual to accompany this (see Report Supplementary Materials 2 and 3), we delivered three groups using the 2.5-day intervention format described. Groups were run between July and September 2016. We approached 79 participants from our sampling frame (see Appendix 1). 14 Thirteen participants attended: six attended the first group, three the second group and four the third group. Difficulty with attendance was as a result of commitment of time over 2 consecutive days.

As part of the formative evaluation the process evaluation team observed group two and concluded that the intervention had been delivered as per protocol. They interviewed 12 out of the 13 participants who attended these feasibility groups. Some participants had also completed a feedback form at the end of each day.

Overall, everyone had appreciated connecting with others with the same condition, and this was a driver to return. The course content and pace were well received. Group discussions were appreciated, as was the lay facilitation. The nurse one-to-one sessions were highly valued by most, with the majority wanting telephone follow-ups. The majority said that they would recommend the course to others. Only two had some reservations: one felt that it would be useful for people newly diagnosed and one wanted more individual tailoring of advice.

We interviewed three nurses and two lay facilitators to ask about their experiences of running the groups. This feedback resulted in the removal of the half-day follow-up because participants found it difficult to get the time off to attend. The final and the third feasibility group piloted the 2-day revised format intervention, which was subsequently adopted for the main RCT (see Figure 4). For the main study we also provided additional facilitator training on medication use.

FIGURE 4.

Intervention structure. (a) Feasibility phase and (b) main study.

During the feasibility phase it became clear that recruiting laypeople with chronic headaches as facilitators would be challenging because of the unpredictable nature of the condition. This was discussed at a trial management meeting in June 2016, at which there was agreement to recruit allied health professionals to co-facilitate with the nurse.

The theoretical underpinnings and behaviour change rational and techniques for the CHESS intervention, and course content, are described elsewhere (see Appendix 4). 19 The CHESS intervention materials are available from http://wrap.warwick.ac.uk/171671.

Choice of clinical effectiveness and cost-effectiveness outcome measures (work package 4)

Objective: to select most appropriate outcome measures for the RCT of the CHESS intervention package.

This work is described in our published papers and appendices (see Appendices 8–10). 26,27

Multicentre trial (work package 5)

Objective: to run a multicentre RCT, including an economic evaluation, of the CHESS intervention package.

Details of the trial have been published elsewhere (see Appendices 14 and 15). 43,44 Further information is available in our original application, final protocol, data management plan, statistics analysis plan and health economics analysis plan (see Report Supplementary Material 5–8). Here we address the research question:

1. Is the CHESS intervention package clinically effective and cost-effective when compared with a usual-care control?

Clinical methods

Practices identified people who had consulted with headaches or who had been prescribed a migraine-specific drug (triptans/pizotifen) in the previous 2 years. 14 The list of people was screened by a GP in the practice to identify people it would not be appropriate to approach, for example people with a severe uncontrolled mental health problem or a terminal illness. Table 1 provides a list of the full inclusion/exclusion criteria. The practice then sent out packs inviting people to express an interest in the trial. Those who were interested returned an expression of interest form to the research team. The research team telephoned the potential participant to confirm their eligibility and obtain verbal consent to start the smartphone app, or paper headache symptom diary. A study pack with the participant information leaflet, consent form and baseline questionnaire was sent to the potential participant. Once the consent form and baseline questionnaire were received, a classification interview call was arranged (see Figure 5).

TABLE 1 - Inclusion and exclusion criteria

Notes

During the lifetime of the study we added living at the same address as someone already randomised to the trial as an exclusion.

Reproduced with permission from Patel et al. 43 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/. The table includes minor additions and formatting changes to the original table.

Copyright © 2023 Underwood et al.

Inclusion criteria

Exclusion criteria

Able and willing to comply with the study procedures and provide written informed consentAged ≥18 years (no upper limit)Living with chronic headache: defined as headache on ≥ 15 or more days per month for at least the preceding 3 monthsThe nurse telephone classification interview confirms headache type to be chronic migraine, or chronic tension-type headache, with or without medication overuse headacheFluent in written and spoken English

Unable to attend the group sessionsNo access to a telephone (for classification interview)Has an underlying serious psychological disorder with ongoing symptoms that preclude or significantly interfere with participation in the group interventionPrevious entry or randomisation in the present trialCurrently participating in another clinical trial of headache treatments or unregistered medicinal product, or < 90 days have passed since completing participation in such a trial

FIGURE 5.

Study flow chart. NMC, National Migraine Centre. Adapted with permission from Patel et al. 43 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

Copyright © 2023 Underwood et al.

Nurses used the classification tool (see Figure 3) to confirm study eligibility and flagged any people with suspected non-eligible headaches for a second telephone interview with a doctor from the NMC. Participants classified as having an eligible headache type were then eligible for randomisation. If there was not a group an individual could attend, they were not randomised.

Our population of interest was people meeting an epidemiological definition of chronic headaches, that is, people with headaches for ≥ 15 days per month for at least 3 months. 8,10,11 At our classification day, it was decided to focus the trial just on people with migraine or tension-type headaches. People suspected of having other chronic headache types were directed to their GPs. For reporting we present three primary headache phenotypes:

1. definite chronic migraine – people meeting International Classification of Headache Disorder, third edition (ICHD-3) criteria for chronic migraine, that is at least 8 days per month with a migraine attack with or without aura9

2. chronic tension type headache and episodic migraine – people meeting ICHD-3 chronic tension-type headache and ICHD-3 criteria for episodic migraine

3. chronic tension-type headache.

Each group included those with and without medication overuse headache. Because not all migraine attacks meet the strict criteria for a migraine attack either because of early treatment or because they are mild,9 we report chronic migraine and chronic tension-type headache with episodic migraine together for our primary analysis. People meeting our definition of probable chronic migraine are typically managed in the same manner as people with definite chronic migraine and received the same advice within the trial.

In the feasibility study, 97 out of 103 (94%) of those assessed as having an eligible headache type had ‘chronic migraine’. 16 With the agreement of the Trial Steering Committee, the Data Monitoring Committee and the funder, we specified that, if at least 85% of our participants had ‘migraine’, that our primary analysis would just be on the population with ‘migraine’, with sample size inflated, if necessary, to ensure adequate statistical power for this analysis.

Between April 2016 and March 2018, we trained 30 facilitators. Quality assurance for the intervention was monitored in several ways, including observations of sessions, audio-recordings of sessions, participant feedback and facilitators’ personal reflections (see Report Supplementary Material 9).

The control intervention quality was assessed by keeping record of when packs had been sent to participants, any contact with control participants, attrition rates, when letters had been sent to the GP and date of classification interview.

Randomisation and masking

The unit of randomisation for this parallel-group study was the individual participant. The randomisation allocation ratio was 1 : 1.07 in favour of the intervention group to account for clustering in one arm. Randomisation was done using minimisation, stratifying by geographical locality (Midlands and Greater London) and headache type (chronic migraine, chronic migraine with episodic migraine or chronic tension-type headache only, with or without medication overuse headache). To maintain allocation concealment all baseline data were collected prior to randomisation. Randomisation was done using Warwick Clinical Trials Unit’s randomisation programme by a person independent of the research team. It was not possible to mask the study team, facilitators or participants from the treatment allocation.

Groups of four or five geographically proximate practices were clustered with the aim of starting recruitment around the same time in several practices. Participants were randomised in batches, with a target size of around 20 to ensure sufficient participants to populate a group and reduce any delay between randomisation and starting the intervention group. Participants unable to attend the group that they were originally allocated to were offered attendance at another group if available.

Participants were informed of their allocation by the study team via a telephone call. Written confirmation of the randomisation and headache classification was also sent to the participant and to their GP.

Post-randomisation withdrawals and exclusions

All participants were followed up when possible, and data were collected in accordance with the trial protocol until the end of the trial. No further data were collected for participants who explicitly withdrew their consent, and only the data collected up the point of withdrawal were used in the final analysis.

Outcome measures

Primary outcome measure

Our primary outcome was the HIT-6 score at 1 year. 33 It consists of six questions with five responses (never to always: score 6, 8, 10, 11 and 13 points). The score ranges from 36 to 78 points, with higher scores indicating greater headache severity (see Report Supplementary Material 10).

Secondary outcome measures

We used the CHQLQ as a secondary headache disability outcome. Our other secondary outcome measures were headache days in the preceding 28 days; typical headache duration and severity in previous 28 days; EuroQoL-5 Dimensions, five-level version (EQ-5D-5L);45 SF-12 version 2,46 reported as physical and mental component scores; Hospital Anxiety and Depression Scale (HADS);47 Pain Self-Efficacy Questionnaire (PSEQ);48 and Social activity: Social Integration Subscale (SIS) of the Health Education Impact Questionnaire (heiQ). 49

Baseline data collected included basic demographic data and data on the troublesomeness of other bodily pains. 50

We collected data on total headache days, average duration of headache and headache severity from participants weekly for 6 months and then monthly, starting from the initial eligibility call. These outcomes were collected according to patient preference using a smartphone app or paper diary records (not both).

We sent postal questionnaires at 4, 8 and 12 months. HIT-6 scores, headache days and EQ-5D-5L scores were, if needed, collected by telephone. To maximise follow-up rates, we used several strategies, including sending high street vouchers with each initial questionnaire and study pens with reminder questionnaires and a shorter questionnaire being sent as a second reminder (see Report Supplementary Material 11).

Sample size

We estimated the sample size using Moerbeek and Wong’s51 method, which accounts for clustering in one arm. Based on similar trials52 we used an intracluster correlation coefficient of 0.01.

The sample size (n = 689: relaxation arm, n = 333; self-management arm, n = 356) was estimated to assess the clinical effectiveness in the migraine population, providing 90% power to detect a target (worthwhile) between-group difference of 2 points (SD 6.87, from the feasibility study) in the HIT-6 outcome at 12 months using a two-sided test and a 5% significance level with 20% loss to follow-up. 14 Some support for this being a plausible effect size came from a pilot study53 of a similar intervention for migraine. Some overrun on sample size was expected to allow all groups to be adequately populated.

Primary and secondary analyses

The primary analysis approach was intention to treat. Data were summarised and reported in accordance with Consolidated Standards of Reporting Trials (CONSORT) guidelines for RCTs. 54 Our statistical analysis plan is available in Report Supplementary Material 7.

Participant characteristics and outcomes were summarised as mean and SD for continuous data or frequency and percentage for categorical data, summarised by treatment arm. The median and IQR were presented if data deviated substantially from a normal distribution.

The primary end point was 12 months. For the primary and secondary analyses, treatment effects were estimated using linear mixed-effects models with partial clustering to account for the trial design with clustering in the self-management arm [command ‘mixed’ from Stata^® (StataCorp LP, College Station, TX, USA)]. Analyses were adjusted for age, gender, the baseline value of the dependent variable and baseline stratification factors (type of headache and geographical locality). The adjusted treatment effect estimates and associated 95% CI s were presented for all analyses. All statistical tests were two-sided at the 5% significance level. Analyses were conducted using the statistical software package Stata 15 and R (The R Foundation for Statistical Computing, Vienna, Austria) version 4.0.3.

Complier-average causal effect analyses

We carried out complier-average causal effect (CACE) analyses for both levels of adherence for the primary outcome only. Minimal adherence was defined as attending day 1 of the intervention plus the one-to-one session. Full adherence was defined as the participant attending both days, plus individualised contact with the nurse.

Subgroup analyses

We carried out prespecified subgroup analyses using formal statistical tests for interaction to examine whether baseline anxiety (HADS anxiety score, ≤ 10 and > 10 points), depression (HADS depression score, ≤10 and >10 points) and severity (HIT-6 score, ≤ 64 and > 64 points) moderated treatment effect. 55

Symptom diary (total headache days, average duration of headache and headache severity)

These data were analysed using longitudinal analyses adjusting for the same variables as those used in the primary analyses (fixed effects) and participant as random effects.

Additional analyses

We assessed treatment effects in terms of the primary outcome for the whole population, including those with tension-type headache only, for chronic migraine and chronic tension-type headache with episodic migraine separately, and for those with or without medication overuse headache. These results will contribute to future meta-analyses and inform future guidelines.

Sensitivity analyses

We performed two sensitivity analyses: one excluding those interviewed for the process evaluation and another excluding those who reported <15 headache days on their baseline questionnaire.

Adverse events and serious adverse events

The frequency and percentage of adverse events (AEs) and serious adverse events (SAEs) in the trial are reported.

Clinical results

The results are in Appendix 15. 44 A detailed statistical report is available as Report Supplementary Material 12. A separate statistical report is also available for all 736 randomised participants, including those without migraine, as Report Supplementary Material 13.

Screening and recruitment

Between April 2017 and March 2019 we recruited from 164 general practices with a combined patient population of 1,529,684. Of the 32,998 potential participants identified from screening we approached 31,020 (94%). We received 2179 expressions of interest (including 41 self-referrals). We contacted 1871 (86%); of these, 1159 (62%) were eligible. Of these, 92 (8%) did not proceed because there were no suitable groups for them to attend and 785 of the remaining 1067 (74%) returned consent forms. We did classification calls with 751 (96%) of these. Nine people (1%) withdrew at this time. Six (<1%) were excluded because of a non-eligible headache (two cluster headache, two new daily persistent headache, one cervicogenic headache and one hemicrania continua; Figure 6). The final number of recruited participants was therefore 736 (including 27 self-referrals).

FIGURE 6.

Consolidated Standards of Reporting Trials flow chart. a, Participants withdrew from intervention only and continued to be on follow-up; b, Complete withdrawal. One withdrawn on the day of randomisation, and one 2 days after randomisation; c, cumulative number of complete withdrawals. Adapted with permission from Underwood et al. 44 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

Copyright © 2023 Underwood et al.

Of these, 727 (99%) had chronic migraine or chronic tension type headache and episodic migraine. Henceforth, unless otherwise specified, all the results refer to this group. Two participants were withdrawn from the trial by the study team soon after randomisation; one because they were living with someone already randomised to the trial, the other because the person made known that they had recently started in a trial of a calcitonin gene-related peptide (CGRP) monoclonal antibody for their headaches.

Participant baseline characteristics

Baseline characteristics were well matched between treatment arms. Most participants were white (586/727; 81%) and the majority were female (604/727; 83%) and the mean age was 48 years (SD 15 years). Just over half (396/727; 55%) had definite chronic migraine, 46% (331/727) had chronic migraine or chronic tension type headache with episodic migraine and 407 out of 727 (56%) had medication overuse headache. The median number of headache/migraine days over the last 4 weeks reported at baseline was 16 (IQR 11–20) days. Thirty-eight per cent (274/721) reported <15 headache days in the preceding 4 weeks.

We reached a diverse population that was representative of national averages in terms of ethnic mix and levels of deprivation, with a good mix of rural and urban areas.

Medication use was similar across both groups. At baseline, 662 out of 727 (91%) had used acute treatments, and 235 out of 727 (32%) had used prophylactic medications.

The overall mean HIT-6 score (primary outcome) at baseline was 64.5 points (scale range 36–78 points; SD 5.5 points), suggesting that most participants had severe symptoms. People with chronic migraine had greater headache severity, lower quality of life, less self-efficacy and less social interaction than those with chronic tension type headache and episodic migraine. Many participants had chronic pain other than their headaches: 375 out of 727 (52%) had at least moderately troublesome neck and 277 out of 727 (38%) had at least moderately troublesome back pain (see Report Supplementary Material 12).

Participant follow-up

Follow-up rates for the primary outcome were 76% (551/727) at 4 months, 75% (546/727) at 8 months and 80% (582/727) at 12 months. Three participants had missing baseline data. The primary analysis included data from 579 out of 727 (80%) participants. Thirty-two (4%) participants withdrew completely from the study including follow-up.

Adverse events

There were two SAEs, both deaths unrelated to the trial. There were seven AEs: five in the self-management arm that occurred during the intervention sessions and were related to developing a migraine or becoming upset during a session. Two were in the usual-care arm: one related to the content of the relaxation CD, and one participant became upset during a process evaluation interview.

Intervention data

We had 42 intervention groups, run by 20 facilitators, at 35 venues in the Midlands and London. Median group size at randomisation was nine (IQR 7–12) and median attendance on day 1 was 6.5 (IQR 5–9). The first session was attended by 286 out of 376 (76%) of those randomised, 259 out of 376 (69%) achieved partial adherence and 216 out of 376 (57%) full adherence (see Report Supplementary Material 12).

Effect of COVID-19

One-year follow-up was due to be completed soon after the UK national lockdown on 23 March 2020. Inability to access the office during this time meant that we were unable to manage reminder questionnaires. For this reason, more 12-month core outcomes were collected by telephone. The COVID-19 pandemic also made it impossible for the study team to visit general practices to collect data for the health economics analyses. We developed new processes to allow practice teams to extract these data remotely on our behalf.

Primary outcome: six-item Headache Impact Test

We found no evidence of a positive effect at 12 months, the primary end point (mean difference –0.3 points, 95% CI –1.23 to 0.67 points), or at 8 months (mean difference 0.07 points, 95% CI –0.95 to 1.09 points) (see Table 2). At 4 months participants in the self-management support group had statistically significantly lower HIT-6 scores (better headache-related quality of life: mean difference of –1.0 point; 95% CI –1.91 to –0.006 points) than participants in the standard-care group. The intracluster correlation coefficients were very small (< 0.0001) in all our analyses. The findings from our CACE analyses, sensitivity analyses and additional analyses were not materially different. There was no evidence of treatment effect modification in our prespecified subgroup analyses.

TABLE 2 - The HIT-6 adjusted treatment differences at different time pointsa

ITT, intention to treat.

Adjusted for age, sex, headache type, geographical locality. and baseline measure of the outcome. Positive difference favours control.

Note

HIT-6 (range 36–78 points; higher = worse).

Adapted with permission from Underwood et al. 44 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

Copyright © 2023 Underwood et al.

Model	Time point (months)
	4	8	12
ITT
Mean difference (95% CI)	–1.0 (–1.91 to -0.006)	0.07 (–0.95 to 1.09)	–0.3 (–1.23 to 0.67)
p-value	0.049	0.888	0.560
CACE (minimum adherence)
Mean difference (95% CI)	–1.3 (–2.57 to -0.02)	0.04 (–1.22 to 1.31)	–0.4 (–1.67 to 0.87)
p-value	0.046	0.945	0.540
CACE (full adherence)
Mean difference (95% CI)	–1.6 (–3.10 to -0.01)	0.05 (–1.46 to 1.56)	–0.5 (–2.00 to 1.05)
p-value	0.048	0.945	0.540

Secondary outcomes

There was no evidence of a difference in headache days between the groups at the 8-month and 12-month follow-ups. However, at 4 months participants in the self-management support group reported 1.5 (95% CI 0.48 to 2.56) more headache days in the previous 28 days than participants in the standard-care group. The mean difference of EQ-5D visual analogue scale score at 12 months was 3.9 points (95% CI 0.90 to 6.88 points), favouring the self-management support group, but there were no statistically significant differences at 4 or 8 months. At 4 and 12 months, but not at 8 months, those in the self-management group had stronger self-efficacy beliefs as measured by the PSEQ. The mean differences were 2.3 points (95% CI, 0.51 to 4.00 points) and 2.1 points (95% CI 0.17 to 3.96 points) at 4 and 12 months, respectively. We did not find any differences in the role restrictions, limitations or emotional impact domains of the CHQLQ, EQ-5D-5L, the SF-12 mental and physical component scores, the HADS, or the SIS of the heiQ at any time point (see Figure 7).

FIGURE 7.

Treatment differences and 95% CI s for secondary outcomes adjusted for age, gender, baseline value of the dependent variable, headache type and geographical locality at the 4-, 8- and 12-month follow-ups. MCS, mental component score; PCS, physical component score; SF-12/36, Short Form questionnaire-12/36 items. Adapted with permission from Underwood et al. 44 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

Copyright © 2023 Underwood et al.

At 8 months the mean headache duration in the self-management group was 9.2 hours (SD 7.3 hours), whereas the mean duration in the standard-care group was 8.0 hours (SD 6.9 hours) (difference, 2.0 hours; 95% CI 0.55 to 3.42 hours). There were no statistically significant differences at 4 or at 12 months (see Report Supplementary Material 12).

Estimates and 95% CI rescaled to range from 0 to 100 for graphical representation purposes only (see Figure 7). To obtain the estimated difference and its 95% CI in its original scale, the value from graph is multiplied by (maximum value/100). For example, the estimated difference for the HADS anxiety score at the 4-month follow-up was (−0.801 × 21/100) −0.16821. Full details of results are available from Appendix 15 and Report Supplementary Material 12).

Medication use

At baseline, 91% of participants were using acute treatments (painkillers/triptans) and 32% were using preventative medications. There was no change over time or any between-group differences in proportions using acute prescribed and over-the-counter acute medications or prophylactic medications. Neither were there any statistically significant between-group differences in the defined daily doses used by those using acute and prophylactic drug treatments.

There were two statistically significant differences in drug use in those reporting use of that group of drugs; more defined daily doses of beta-blockers (p = 0.005) at 4 months and fewer defined daily doses of opioids (p = 0.02) were used at 8 months in the self-management support group (see Appendix 15 and Report Supplementary Material 12).

There were no differences in proportions using acute medications for ≥10 or ≥15 days in the last 28 days at any time point. Overall, at 12 months, 43%, and 21% participants, respectively, used painkillers/triptan for headaches on ≥10 or ≥15 days out of the last 28 days. This compares with 63% and 38%, respectively, at baseline (see Report Supplementary Material 12).

Second-line prophylactic drugs [Botox^® (AbbVie Inc., North Chicago, IL, USA) and CGRP monoclonal antibodies] were used by five participants. Four received Botox injection (self-management support group, n = 2; standard-care group, n =2) and two people from the self-management support group were prescribed erenumab. One participant received both Botox and erenumab.

Headache symptom diary

No statistically significant differences were observed between the two groups from the longitudinal analyses for any of these outcomes. The estimated between-group difference for the number of headache days over 1 year was 0.2 (95% CI –0.11 to 0.46) days, for the duration of headache the estimated difference was 0.4 (95% CI –0.47 to 1.28) hours and for headache severity the estimated difference was 0.2 (95% CI –0.08 to 0.46) points on a 0–11 scale.

Methods

We used a mixed-methods approach. Quantitatively, we described reach/context, recruitment, dose delivered, dose received and fidelity. Qualitatively, we explored the experiences of participants, intervention facilitators and GPs about their involvement in the trial (see Table 4).

Results

Conclusion of process evaluation

The process evaluation suggests that CHESS reached a diverse population across different geographical settings. Attendance reached our predefined dose, but many participants were not exposed to the intervention. The interventions were delivered with fidelity and, although generally well received, with some sessions liked more than others, both intervention facilitators and participants had reservations about several components of the course. There is evidence that participants valued the group and one-to-one aspects of the intervention giving them the opportunity to explore and review their headaches and its management. The process evaluation provides no clear explanation as to why the CHESS intervention appears ineffective.

Development of research idea

The original research question, ‘Will an education and self-management programme help people living with chronic headache?’, is based on a research recommendation in the 2012 NICE guidelines on the management of headaches. 5 Members of the lay guideline development group helped identify the research question.

Grant application

Our co-applicants included representatives of the three leading UK migraine charities. The involvement at this stage was from our charity partners. It did not include direct patient involvement.

Establishing a lay advisory group

We established a lay advisory group, from our larger lay reference group, at the start of the programme. We sought to establish a group of 8–10 people living with chronic headaches. Members needed access to e-mail but not specific skills. Our charity partners approached a diverse group of people with chronic headaches. We also approached people through a University of Warwick PPI initiative. We contacted responders to identify their headache type and basic demographic details. Ten people joined the CHESS lay advisory group. This group then developed their own rules of engagement (see Appendix 19, box 2). 71

Trial oversight

We invited representatives of each of the three charities to our monthly trial management groups. The charities found it difficult to commit their staff to meeting times. Our independent PSC had two lay members who contributed to eight meetings over the duration of the programme. One lay representative contributed to our Data Monitoring Committee.

Feasibility study research ethics application

When we sought research ethics approval for the feasibility study our lay advisory group was not established. Our charity partners helped us to find five people who critically reviewed our patient-facing documents.

Headache classification day

Our lay reference group contributed to the classification day, ensuring that the key questions generated were relevant to people with chronic headache and health professionals. Seven lay people attended on the day.

Electronic data collection

The CHESS lay advisory group helped to refine the questions used in the electronic diary and provided feedback on the acceptability, the user experience and the app. A suggestion from our lay advisory group included the addition of a calendar indicating the period of recall.

We provided participants with a summary of their diary data once data collection was complete. The group commented on, and made recommendations to improve, the individual data summary sent to main trial participants at the end of their data collection period.

Intervention design and development

Seven members of Migraine Action took part in an individual, face-to-face interview about their headaches and the treatments they had tried, to inform the intervention development.

Our intervention development day was chaired by one of our professional charity partners and attended by three lay representatives, suggested by our charity partners.

Patient-reported outcome measures

During the feasibility study three members of the lay advisory group helped analyse data from interviews exploring participants’ views on the relevance, clarity, comprehensiveness and acceptability of four PROMs.

Core outcome set for migraine

Three PPI members actively collaborated in all stages of the core outcome set for migraine (COSMIG) project, allowing co-development of a three-stage international Delphi survey, subsequent data interpretation, co-facilitation of the consensus meeting and co-authorship of the subsequent publication. 39

Public and media engagement

Following a university press release we were invited onto The Victoria Derbyshire Show (BBC News, London, UK). We were represented by one of our lay members and the chief investigator. Our lay member was able to speak about their experience of chronic headaches and the impact this had on their life. This generated public interest in CHESS.

Post-results patient and public (lay) discussion group

Once the trial outcome was known we sought an additional patient and public perspective on the findings. We ran a 2-hour virtual discussion group with 10 people drawn from the CHESS reference group. The process evaluation team presented an overview of the process evaluation results followed by the trial outcomes. The group were disappointed with the outcome but not completely surprised. Although the CHESS intervention provided useful knowledge and tools, the group thought that developing new skills and behaviours takes time. Participants thought a longer intervention may be needed to reinforce learning and to support behaviour change.

Patient and public involvement learning points

Our thinking on patient involvement has developed substantially since we submitted the first application for funding in June 2012. At that time, we had planned a formal evaluation of PPI in the programme. We were asked to remove this by the funding board at the outline stage. This does mean we have missed the opportunity to do a more in-depth prospective evaluation. Nevertheless, we have identified some key learning points:

1. More training in the research process for both our lay and professional PPI partners.

2. Lay input, in addition to charity input, into the development of the grant may have provided a different perspective.

3. A designated team member to be the PPI contact could have improved engagement.

4. Regular correspondence to update the lay advisory group throughout the duration of the programme.

Reflections on what was and what was not successful in the programme

Overall, the programme was delivered well, and we achieved our overarching aim of developing and testing a self-management support programme for people living with chronic headaches. We successfully undertook a series of systematic reviews to set the scene for the rest of the work. Our feasibility study was complicated as we were seeking to achieve multiple objectives within the same framework. This did make it difficult to explain to participating general practices and study participants. Recruitment rates for the main trial from general practice were smaller than originally thought. In our original application we thought we would need to work with 30–40 general practices. In the end we needed to work with 164 practices.

Limitations relating to the method or execution of the research

This was an ambitious programme of work that in phase 1 needed integration of interlinked work packages. The original timelines were, in some areas, too ambitious, needing some changes as to how we collected data, notably recruiting for some interviews through our charity partners rather than from feasibility study participants. In addition, work on validating the CHQLQ was not completed prior to starting the main trial, meaning that we set the HIT-6 as our primary outcome, although our view now is that this is not the most suitable outcome for studies of this nature. This has not, however, affected our final conclusion as the results from the two measures are consistent. The COVID-19 pandemic had a limited impact on the trial. Follow-up finished early in the first wave of the pandemic, meaning that we had some reduction in the data available for our secondary outcomes and we needed to make changes to how we collected data from GP surgeries.

Conclusions from the whole programme

Our data effectively exclude the possibility that this short intervention is effective for the treatment of people with chronic migraines, or chronic tension-type headache and episodic migraine. Nevertheless, the health burden of chronic headache disorders, principally chronic migraine, is debilitating. Further advances in this field must be driven by new theoretically and/or biologically informed intervention models.

Recommendations for future research

1. New work to better understand the health impact of chronic headache disorders and to identify modifiable risk factors for a poor outcome.

2. Development and testing of new non-pharmacological interventions for a tightly phenotyped group with chronic migraine.

3. Work is needed to improve classification of headache disorders in primary care to allow better targeting of the available drug treatments of proven effectiveness, and reduce medication overuse.

Implications for practice and any lessons learnt

There is no need, at this time, to implement formal supportive self-management support programmes, of the type tested here, for people living with chronic migraine.

Trial team

1. Felix Achana (Associate Professor).

2. Dawn Carnes [Director: National Council for Osteopathic Research (UK)].

3. Sandra Eldridge (Professor).

4. David Ellard (Associate Professor).

5. Barbara Gougoulis (Administrator).

6. Frances Griffiths (Professor of Medicine in Society).

7. Elizabeth Habershon (Research Nurse).

8. Kirstie Haywood (Reader).

9. Siew Wan Hee (Senior Research Fellow).

10. Helen Higgins (Senior Project Manager).

11. Sobhash Jhuree (Research Nurse).

12. Manjit Matharu (Professor).

13. Hema Mistry (Associate Professor).

14. Dipesh Mistry (Trial Statistician).

15. Sian Newton (Trial Manager).

16. Vivien Nichols (Research Fellow).

17. Chloe Norman (Trial Co-ordinator).

18. Emma Padfield (Senior Project Manager).

19. Shilpa Patel (Senior Research Fellow).

20. Stavros Petrou (Professor).

21. Tamar Pincus (Professor).

22. Rachel Potter (Senior Research Fellow).

23. Harbinder Sandhu (Professor).

24. Kimberley Stewart (Trial Manager).

25. Stephanie Taylor (Professor).

26. Martin Underwood (Professor).

Trial Steering Committee

1. David Fitzmaurice.

2. Simon Kyle.

3. Louise Longworth.

4. Andrea Manca.

5. Stephen Morley.

6. Allison Ogden-Newton.

7. Fran Reynolds.

8. Leone Ridsdale (chairperson).

9. Alexandra Sinclair.

10. Linda Williams.

11. Nicola Williams.

Data Monitoring and Ethics Committee

1. Brian Burns.

2. Professor Lee Shepston (chairperson).

3. Tracey Young.

Grant-holding body

Ceri Jones, Giovanni Bucci, and the finance team within University Hospitals Coventry and Warwickshire.

Support staff

Classification day

Angela Abdalla, Maha Ahmed, Debi Alvares, Jane Anderson, Bal Athwal, Mohammed Belhag, Barbara Box, Phoebe Brobby, Sam Chong, Katie Dodd, Paul Dorman, Julie Edwards, Roland Etti, Jithin George, Rachel Hanson, Susan Hirst, Graham Howkins, Sarah Irving, Zaza Katsarava, Francesca Kelly, Susie Lagrata, Nassif Mansour, Lynne Muldoon, Gemma Napul, Niran Nirmalananthan, Sally Reckert, Deb Smith, Imad Soryal, Tim Steiner, Anthony Thomas, Alok Tyagi, Jitka Vanderpol, Ben Wakerley, Tom Walker, Stuart Weatherby, Lara York and Tim Young.

Core Outcome Set in Migraine consensus day

Maha Ahmed, Bal Athwal, Barbara Box, Keith Griffiths, Asha Hareendran, Amanda Harrison, Kirstie Haywood, Rigor Jensen, Jessica Jones, Susie Lagrata, Richard Lipton, Anne-Marie Logan, Manjit Matharu, Vivien Nichols, Niranjanan Nirmalananthan, Shilpa Patel, Gemma Pearce, Stavros Petrou, Rachel Potter, Helen Power, Regina Rendas-Baum, Susan Stubbs, Martin Underwood, Amit Vora and Lara York.

Clinvivo

Johnathan Foss and Rob Froud.

Additional authors of publications

Fiona Caldwell, Tom Mars and Arani Vivekanantham.

Classification calls

Tracey Adcock, Elaine Butcher, Jon Davies, Pauline Derbyshire, Linda Field, Eleanor Hoverd, Sarah Joshi, Claire Talbot, Sobhash Jhuree and Elizabeth Habershon.

Lay advisory group members

Barbara Box, Sarah Alix, Graham Howkins, Lynne Muldoon, Gemma Pearce, Sally Reckert, Deb Smith, Lisa Stubbs, Christine Williams and Lara York.

Lay contributors to feedback on feasibility Research Ethics Committee application

Rachel Clegg, Claire Horton, Fran Kelly and Sheila Kelly.

Lay contributors to intervention design day

Elinor Bailey, Andrew Cook and Meghan Fay.

Facilitators

Jeanette Allison, Sue Carrington, Olivia Neely, Emma Randle and Gemma Pearce. Denis Anthony, Melissa Baldey, Claire Balmer, Sarah Bridgewater, Sue Carrington, Sinead Clarke O’Neill, Penny Crofts, Rebecca Davies, Linda Field, Elizabeth Habershon, Joanna Haywood, Sobhash Jhuree, Debbie Kelly, Sue Kenney, Anne-Marie Logan, Janet Lowe, Vinod Mahtani, Anna Molares, Katherine Priddis, Katrin Probyn, Jayshireen Singh, Karlene Stimpson, Priya Varma, Claire Winch and Sarah Wytrykowski.

General practices

Clinical Commissioning Groups

1. Coventry and Rugby.

2. South Warwickshire.

3. North Warwickshire.

4. Birmingham Cross City.

5. East Staffordshire.

6. South East Staffordshire and Seisdon Peninsula.

7. Birmingham South Central.

8. Nottingham City.

9. Nottingham North & East.

10. Nottingham West.

11. Leicester City.

12. East Leicester & Rutland.

13. Camden.

14. Hounslow.

CHESS digital versatile disc

Slate & Mortar Ltd (Birmingham, UK).

CHESS mindfulness compact disc and handout

Mindlab Goodworks Ltd (Teddington, UK) and Gill Davies.

Relaxation compact disc production

Warwick Audio Visual Department.

National Migraine Centre

Paul Booton, David Bloomfield, Jessica Briscoe, Charlotte Burr, Nazeli Manukyan, Katy Munro, Judith Pearson, Swati Raina, Lauren Shirazi and Heather Sims.

Migraine Trust

Angus Baldwin, Susan Haydon, Wendy Thomas and Arlene Wilkie.

Migraine Action

Simon Evans.

Clinical Research Networks

1. Leicester.

2. North London.

3. Nottingham.

4. South London.

5. Thames Valley South Midlands.

6. West Midlands.

Contributions of authors

Martin Underwood (https://orcid.org/0000-0002-0309-1708) was the chief investigator for the programme. Martin Underwood and Manjit Matharu (https://orcid.org/0000-0002-4960-2294) devised the original concept and Manjit Matharu was clinical lead. Dawn Carnes (https://orcid.org/0000-0002-3152-3133), Sandra Eldridge (https://orcid.org/0000-0001-5638-2317), David R Ellard (https://orcid.org/0000-0002-2992-048X), Frances Griffiths (https://orcid.org/0000-0002-4173-1438), Kirstie Haywood (https://orcid.org/0000-0002-5405-187X), Siew Wan Hee (https://orcid.org/0000-0002-0415-263X), Harbinder Sandhu (https://orcid.org/0000-0003-1522-8078), Stavros Petrou (https://orcid.org/0000-0003-3121-6050), Tamar Pincus (https://orcid.org/0000-0002-3172-5624), Stephanie JC Taylor (https://orcid.org/0000-0001-7454-6354) and Manjit Matharu were the original co-applicants for the programme grant.

Sian Newton (https://orcid.org/0000-0001-8035-0411), Choe Norman (https://orcid.org/0000-0002-3275-3141) and Kimberley Stewart (https://orcid.org/0000-0002-9814-8331) were trial managers and Helen Higgins (https://orcid.org/0000-0002-7095-4542) and Emma Padfield (https://orcid.org/0000-0003-4832-9609) were senior project managers. Sandra Eldridge, Siew Wan Hee and Dipesh Mistry (https://orcid.org/0000-0002-0875-9260) led the statistical analysis and interpretation. Felix Achana (https://orcid.org/0000-0002-8727-9125), Hema Mistry (https://orcid.org/0000-0002-5023-1160) and Stavros Petrou led the health economic analysis and interpretation. Dawn Carnes, Shilpa Patel (https://orcid.org/0000-0003-0726-4888), Harbinder Sandhu, Tamar Pincus and Stephanie JC Taylor and led the development of the self-management intervention, and Rachel Potter (https://orcid.org/0000-0001-6655-8996) and Manjit Matharu led the headache classification and one-to-one consultations. Kirstie Haywood led the outcome measurement work. David R Ellard, Frances Griffiths, Vivien Nichols (https://orcid.org/0000-0002-3372-1395) and Stephanie JC Taylor led the process evaluation.

All authors made substantial contributions to the design of the work or the acquisition, analysis or interpretation of the data, and provided a critical review and final approval of the report.

Publications

Data-sharing statement

All requests for data should be sent to the Warwick Clinical Trials Unit data access team (wctudataaccess@warwick.ac.uk). Access to anonymised data may be granted following review.

Patient data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

Disclaimers

This report presents independent research funded by the National Institute for Health and Care Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, CCF, PGfAR or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, the PGfAR programme or the Department of Health and Social Care.

Background

Self-management support programmes are effective in a range of chronic conditions; however, there is limited evidence for their use in the treatment of chronic headaches. The aim of this study was to test the feasibility of four key aspects of a planned, future evaluative trial of a new education and self-management intervention for people with chronic headache: (1) recruiting people with chronic headache from primary care, (2) a telephone interview for the classification of chronic headaches, (3) the education and self-management intervention itself, and (4) the most appropriate patient-reported outcomes (PROMS).

Methods

Participants were identified and recruited from general practices in the West Midlands region of the UK. We developed a nurse-led chronic headache classification interview and assessed agreement with an interview with headache specialists. We developed and tested a group-based education and self-management intervention to assess training and delivery receipt using observation, facilitator, and participant feedback. We explored the acceptability and relevance of PROMs using postal questionnaires, interviews and a smartphone app.

Results

Fourteen practices took part in the study and participant recruitment equated to 1.0/1000 registered patients. Challenges to recruitment were identified. We did 107 paired headache classification interviews. The level of agreement between nurse and doctor interviews was very good. We piloted the intervention in four groups with 18 participants. Qualitative feedback from participants and facilitators helped refine the intervention including shortening the overall intervention and increasing the facilitator training time. Participants completed 131 baseline questionnaires, measurement data quality, reliability and validity for headache-specific and generic measures were acceptable.

Conclusion

This study indicated that recruiting people with chronic headache from primary care is feasible but challenging, our headache classification interview is fit for purpose, our study intervention is viable, and our choice of outcome measures is acceptable to participants in a future randomised controlled trial (RCT).

Background or aim

Despite guidelines and the International Classification of Headache Disorders (ICHD-III beta) criteria, the diagnosis of common chronic headache disorders can be challenging for non-expert clinicians. The aim of the review was to identify headache classification tools that could be used by a non-expert clinician to classify common chronic disorders in primary care.

Methods

We conducted a systematic literature review of studies validating diagnostic and classification headache tools published between January 1988 and June 2016 from key databases: MEDLINE, ASSIA, Embase, Web of Knowledge and PsycINFO. Quality assessment was assessed using items of the Quality of Diagnostic Accuracy Studies (QUADAS-2).

Results

The search identified 38 papers reporting the validation of 30 tools designed to diagnose, classify or screen for headache disorders: nine for multiple headache types, and 21 for one headache type only. We did not identify a tool validated in primary care that can be used by a non-expert clinician to classify common chronic headache disorders and screen for primary headaches other than migraine and tension-type headache in primary care.

Conclusions

Despite the availability of many headache classification tools we propose the need for a tool that could support primary care clinicians in diagnosing and managing chronic headache disorders within primary care, and allow more targeted referral to headache specialists.

Background

For a trial of supportive self-management for people with chronic headache we needed to develop and validate a telephone classification interview that can be used by a non-headache specialist to classify common chronic headache types in primary care. We aimed to specifically exclude secondary headaches other than medication overuse, exclude primary headache disorders other than migraine and tension-type headache (TTH), distinguish between chronic migraine and chronic TTH, and identify medication overuse headache.

Methods

We held a headache classification consensus conference to draw on evidence and expertise to inform the content of a logic model underpinning the classification interview. Nurses trained to use the logic model did telephone classification interviews with participants recruited from primary care. Doctors specialising in headache did a second validation interview.

Results

Twenty-six delegates attended the headache classification conference including headache specialist doctors, nurses and lay representatives (with chronic headache). We trained six nurses to do the classification interviews and completed 107 paired interviews; median days between interviews was 32 days (interquartile range 21–48 days). We measured level of agreement between the nurse and doctor interviews using proportion of concordance, simple kappa and prevalence-adjusted bias-adjusted kappa (PABAK). Proportion of concordance of agreement between nurse and doctor interviews was 0.76, simple kappa coefficient κ = 0.31 (95% CI 0.09 to 0.52), and PABAK 0.51 (95% CI 0.35 to 0.68), a moderate agreement. In a sensitivity test following review of headache characteristics recorded, concordance was 0.91, κ = 0.53 (95% CI 0.28 to 0.79), and PABAK 0.81 (95% CI 0.70 to 0.92), a very good agreement.

Conclusion

We developed and validated a new evidence-based telephone classification interview that can be used by a non-headache specialist to classify common chronic headache types in primary care.

Background

Self-management interventions are well recognised and widely used in chronic conditions. Their application to chronic headaches has been limited and generally of low quality. We describe here our process for developing an evidence-based, and theory driven, education and self-management intervention for those living with chronic headache.

Methods

Our intervention was designed using several core information sources: the results of three systematic reviews; qualitative material from those living with chronic headaches; our knowledge from existing self-management interventions; and finally collaborative input from a multidisciplinary team of clinicians, academics, patients and charity partners. We manualised the intervention and associated training as a package for use in a feasibility study. We made adaptations for its use in a randomised controlled trial.

Results

We piloted the intervention in four groups with a total of 18 participants. Qualitative feedback from 12 participants and five facilitators allowed the intervention to be refined for the main randomised controlled trial. Some of the key changes included shortening of the overall intervention, changes to the originally planned facilitators and spreading the facilitator training over 3 days rather than 2.

We are now testing the final revised intervention in a randomised controlled trial of its clinical effectiveness and cost effectiveness. The group component of the intervention is delivered over 2 days with the first day focused on living, understanding and dealing with chronic headaches and the second day exploring how to adapt and take control of one’s life with chronic headaches.

Conclusion

Our pilot work indicates that our intervention is feasible to deliver, and with the relevant changes would be acceptable for use with this population. Our randomised control trial is ongoing. We anticipate publishing final results in 2021. The CHESS intervention materials are available from http://wrap.warwick.ac.uk/171671/1/Chronic-Headache-Education-and-Self-management-Support-study-CHESS-additional-information.zip.

Objective

To systematically review the qualitative literature of the lived experience of people with a chronic headache disorder.

Background

Chronic headaches affect 3–4% of the population. The most common chronic headache disorders are chronic migraine, chronic tension-type headache and medication overuse headache. We present a systematic review and meta-ethnographic synthesis of the lived experience of people with chronic headache.

Methods

We searched seven electronic databases, hand-searched nine journals and used a modified Critical Appraisal Skills Programme checklist to appraise study quality. Following thematic analysis we synthesised the data using a meta-ethnographic approach.

Results

We identified 3586 unique citations; full texts were examined for 86 studies and four were included in the review. Included studies differed in their foci: exploring, patient-centred outcomes, chronic headache as a socially invisible disease, psychological processes mediating impaired quality of life and the process of medication overuse. Initial thematic analysis and subsequent synthesis gave three overarching themes: ʻheadache as a driver of behaviour’ (directly and indirectly), ‘the spectre of headache’ and ‘strained relationships’.

Conclusion

This meta-synthesis of published qualitative evidence demonstrates that chronic headaches have a profound effect on people’s lives, showing similarities with other pain conditions. There were insufficient data to explore the similarities and differences between different chronic headache disorders.

Objective

To identify predictors of prognosis and trial outcomes in prospective studies of people with chronic headache.

Methods

This was a systematic review of published literature in peer-reviewed journals. We included (1) randomized controlled trials (RCTs) of interventions for chronic headache that reported subgroup analyses and (2) prospective cohort studies, published in English, since 1980. Participants included adults with chronic headache (including chronic headache, chronic migraine, and chronic tension-type headache with or without medication overuse headache). We searched key databases using free text and MeSH terms. Two reviewers independently extracted data and assessed the methodologic quality of studies and overall quality of evidence identified using appropriate published checklists.

Results

We identified 16,556 titles, removed 663 duplicates, and reviewed 199 articles, of which 27 were included in the review—17 prospective cohorts and 10 RCTs with subgroup analyses reported. There was moderate-quality evidence indicating that depression, anxiety, poor sleep and stress, medication overuse, and poor self-efficacy for managing headaches are potential prognostic factors for poor prognosis and unfavorable outcomes from preventive treatment in chronic headache. There was inconclusive evidence about treatment expectations, age, age at onset, body mass index, employment, and several headache features.

Conclusion

This review identified several potential predictors of poor prognosis and worse outcome postinterventions in people with chronic headache. The majority of these are modifiable. The findings also highlight the need for more longitudinal high-quality research of prognostic factors in chronic headache.

Objectives

To assess the effect of non-pharmacological self-management interventions against usual care, and to explore different components and delivery methods within those interventions.

Participants

People living with migraine and/or tension-type headache.

Interventions

Non-pharmacological educational or psychological self-management interventions, excluding biofeedback and physical therapy. We assessed the overall effectiveness against usual care on headache frequency, pain intensity, mood, headache-related disability, quality of life and medication consumption in meta-analysis. We also provide preliminary evidence on the effectiveness of intervention components and delivery methods.

Results

We found a small overall effect for the superiority of self-management interventions over usual care, with a standardised mean difference (SMD) of –0.36 (–0.45 to –0.26) for pain intensity, –0.32 (–0.42 to –0.22) for headache-related disability, 0.32 (0.20 to 0.45) for quality of life and a moderate effect on mood [SMD 0.53 (–0.66 to –0.40)]. We did not find an effect on headache frequency [SMD –0.07 (–0.22 to 0.08)].

Assessment of components and characteristics suggests a larger effect on pain intensity in interventions that included explicit educational components [–0.51 (–0.68 to –0.34) vs. –0.28 (–0.40 to –0.16)], mindfulness components [–0.50 (–0.82 to –0.18) vs. 0.34 (–0.44 to –0.24)] and in interventions delivered in groups versus one-to-one delivery [0.56 (–0.72 to –0.40) vs. –0.39 (–0.52 to –0.27)] and larger effects on mood in interventions including a cognitive–behavioural therapy (CBT) component with an SMD of –0.72 (–0.93 to –0.51) compared with those without CBT –0.41 (–0.58 to –0.24).

Conclusion

Overall we found that self-management interventions for migraine and tension-type headache are more effective than usual care in reducing pain intensity, mood and headache-related disability, but have no effect on headache frequency. Preliminary findings also suggest that including CBT, mindfulness and educational components in interventions, and delivery in groups, may increase effectiveness.

Aims

To critically appraise, compare and synthesise the quality and acceptability of multi-item patient-reported outcome measures for adults with chronic or episodic headache.

Methods

Systematic literature searches of major databases (1980–2016) to identify published evidence of PROM measurement and practical properties. Data on study quality (COSMIN), measurement and practical properties per measure were extracted and assessed against accepted standards to inform an evidence synthesis.

Results

From 10,903 reviewed abstracts, 103 articles were assessed in full: 46 provided evidence for 23 PROMs, eleven specific to the health-related impact of migraine (n = 5) or headache (n = 6), six assessed migraine-specific treatment response/satisfaction and six were generic measures. Evidence for measurement validity and score interpretation was strongest for two measures of impact, Migraine-Specific Quality of Life Questionnaire (MSQ v2.1) and Headache Impact Test 6-item (HIT-6), and one of treatment response, the Patient Perception of Migraine Questionnaire (PPMQ-R). Evidence of reliability was limited, but acceptable for the HIT-6. Responsiveness was rarely evaluated. Evidence for the remaining measures was limited. Patient involvement was limited and poorly reported.

Conclusion

While evidence is limited, three measures have acceptable evidence of reliability and validity: HIT-6, MSQ v2.1 and PPMQ-R. Only the HIT-6 has acceptable evidence supporting its completion by all ‘headache’ populations.

Objective

To compare the quality and acceptability of a new headache-specific patient-reported measure, the Chronic Headache Quality of Life Questionnaire (CHQLQ) with the six-item Headache Impact Test (HIT-6), in people meeting an epidemiological definition of chronic headaches.

Methods

Participants in the feasibility stage of the Chronic Headache Education and Self-management Study (CHESS) (n = 130) completed measures three times during a 12-week prospective cohort study. Data quality, measurement acceptability, reliability, validity, responsiveness to change and score interpretation were determined. Semistructured cognitive interviews explored measurement relevance, acceptability, clarity and comprehensiveness.

Results

Both measures were well completed with few missing items. The CHQLQ's inclusion of emotional well-being items increased its relevance to participant's experience of chronic headache. End effects were present at item level only for both measures. Structural assessment supported the three and one-factor solutions of the CHQLQ and HIT-6, respectively. Both the CHQLQ (range 0.87 to 0.94) and HIT-6 (0.90) were internally consistent, with acceptable temporal stability over 2 weeks (CHQLQ range 0.74 to 0.80; HIT-6 0.86). Both measures responded to change in headache-specific health at 12 weeks [CHQLQ smallest detectable change (improvement) range 3 to 5; HIT-6 2.1].

Conclusions

While both measures are structurally valid, internally consistent, temporally stable and responsive to change, the CHQLQ has greater relevance to the patient experience of chronic headache.

Background

The Headache Impact Test (HIT-6) and the Chronic Headache Questionnaire (CH-QLQ) measure headache-related quality of life but are not preference-based and therefore cannot be used to generate health utilities for cost-effectiveness analyses. There are currently no established algorithms for mapping between the HIT-6 or CH-QLQ and preference-based health-related quality-of-life measures for the chronic headache population.

Methods

We developed algorithms for generating EQ-5D-5L and SF-6D utilities from the HIT-6 and the CHQLQ using both direct and response mapping approaches. A multi-stage model selection process was used to assess the predictive accuracy of the models. The estimated mapping algorithms were derived to generate UK tariffs and was validated using the Chronic Headache Education and Self-management Study (CHESS) trial dataset.

Results

Several models were developed that reasonably accurately predict health utilities in this context. The best performing model for predicting EQ-5D-5L utility scores from the HIT-6 scores was a Censored Least Absolute Deviations (CLAD) (1) model that only included the HIT-6 score as the covariate (mean squared error (MSE) 0.0550). The selected model for CH-QLQ to EQ-5D-5L was the CLAD (3) model that included CH-QLQ summary scores, age, and gender, squared terms and interaction terms as covariates (MSE 0.0583). The best performing model for predicting SF-6D utility scores from the HIT-6 scores was the CLAD (2) model that included the HIT-6 score and age and gender as covariates (MSE 0.0102). The selected model for CH-QLQ to SF-6D was the OLS (2) model that included CH-QLQ summary scores, age, and gender as covariates (MSE 0.0086).

Conclusion

The developed algorithms enable the estimation of EQ-5D-5L and SF-6D utilities from two headache-specific questionnaires where preference-based health-related quality of life data are missing. However, further work is needed to help define the best approach to measuring health utilities in headache studies.

Objective

Typically, migraine prevention trials focus on reducing migraine days. This narrow focus may not capture all that is important to people with migraine. Inconsistency in outcome selection across trials limits the potential for data pooling and evidence synthesis. In response, we describe the development of core outcome set for migraine (COSMIG).

Design

A two-stage approach sought to achieve international, multistakeholder consensus on both the core domain set and core measurement set. Following construction of a comprehensive list of outcomes, expert panellists (patients, health-care professionals and researchers) completed a three-round electronic-Delphi study to support a reduction and prioritisation of core domains and outcomes. Participants in a consensus meeting finalised the core domains and methods of assessment. All stages were overseen by an international core team, including patient research partners.

Results

There was a good representation of patients [episodic migraine (n = 34) and chronic migraine (n = 42)] and health-care professionals (n = 33) with high response and retention rates. The initial list of domains and outcomes was reduced from >50 to 7 core domains for consideration in the consensus meeting, during which a two-domain core outcome set was agreed.

Conclusion

International and multistakeholder consensus emerged to describe a two-domain core outcome set for reporting research on preventive interventions for chronic and episodic migraine: migraine-specific pain and migraine-specific quality of life. Intensity of migraine pain assessed with an 11-point Numerical Rating Scale and the frequency as the number of headache/migraine days over a specified time period. Migraine-specific quality of life assessed using the Migraine Functional Impact Questionnaire.

Background

To systematically review studies quantifying the association between primary chronic headaches and persistent low back pain (LBP).

Main text

We searched five electronic databases. We included case-control, cross-sectional and cohort studies that included a headache and back pain free group, reporting on any association between persistent LBP and primary headache disorders. Methodological quality was assessed using the Newcastle–Ottawa Scale. Our primary outcome was the association between primary headache disorders and persistent LBP. Our secondary outcomes were any associations between severity of LBP and severity of headache, and the relationship between specific headache sub-types classified as per International Classification of Headache Disorders (ICHD) criteria and persistent LBP.

We included 14 studies. The sizes of the studies ranged from 88 participants to a large international study with 404,206 participants. Odds ratios for the association were between 1.55 [95% confidence interval (CI) 1.13 to 2.11] and 8.00 (95% CI 5.3 to 12.1). Study heterogeneity meant statistical pooling was not possible. Only two studies presented data investigating persistent LBP and chronic headache disorders in accordance with ICDH criteria.

Conclusions

We identified a positive association between persistent LBP and primary headache disorders. The quality of the review findings is limited by diversity of populations, study designs and uncertainty about headache and LBP definitions.

Introduction

Chronic headaches are poorly diagnosed and managed and can be exacerbated by medication overuse. There is insufficient evidence on the non-pharmacological approaches to helping people living with chronic headaches.

Methods and analysis

Chronic Headache Education and Self-management Study is a pragmatic randomised controlled trial to test the effectiveness and cost-effectiveness of a self-management education support programme on top of usual care for patients with chronic headaches against a control of usual care and relaxation. The intervention is a 2-day group course based on education, personal reflection and a cognitive–behavioural approach, plus a nurse-led one-to-one consultation and follow-up over 8 weeks. We aim to recruit 689 participants (356 to the intervention arm and 333 to the control arm) from primary care and self-referral in London and the Midlands. The trial is powered to show a difference of 2.0 points on the Headache Impact Test, a patient-reported outcome measure at 12 months post randomisation. Secondary outcomes include health-related quality of life, self-efficacy, social activation and engagement, anxiety and depression, and health-care utilisation. Outcomes are being measured at 4, 8 and 12 months. Cost-effectiveness will be expressed in terms of incremental cost per quality-adjusted life-year gained.

Ethics and dissemination

This trial will provide data on effectiveness and cost-effectiveness of a self-management support programme for chronic headaches. The results will inform commissioning of services and clinical practice. North West – Greater Manchester East Research Ethics Committee have approved the trial. The current protocol version is 3.6 date 7 March 2019.

Background and objectives

Chronic headache disorders are a major cause of pain and disability. Education and supportive self-management approaches could reduce the burden of headache disability. We tested the effectiveness of a group educational and supportive self-management program for people living with chronic headaches.

Methods

This was a pragmatic randomised controlled trial. Participants were aged 18 years or older with chronic migraine or chronic tension-type headache, with or without medication overuse headache. We primarily recruited from general practices. Participants were assigned to either a 2-day group education and self-management program, a one-to-one nurse interview, and telephone support or to usual care plus relaxation material. The primary outcome was headache related-quality of life using the Headache Impact Test (HIT)-6 at 12 months. The primary analysis used intention-to-treat principles for participants with migraine and both baseline and 12-month HIT-6 data.

Results

Between April 2017 and March 2019, we randomised 736 participants. Because only 9 participants just had tension-type headache, our main analyses were on the 727 participants with migraine. Of them, 376 were allocated to the self-management intervention and 351 to usual care. Data from 586 (81%) participants were analyzed for primary outcome. There was no between-group difference in HIT-6 (adjusted mean difference = −0.3, 95% CI −1.23 to 0.67) or headache days (0.9, 95% CI −0.29 to 2.05) at 12 months. The Chronic Headache Education and Self-management Study intervention generated incremental adjusted costs of £268 (95% CI, £176–377) [USD383 (95% CI USD252 to USD539)] and incremental adjusted quality-adjusted life years (QALYs) of 0.031 (95% CI −0.005 to 0.063). The incremental cost-effectiveness ratio was £8,617 (USD12,322) per QALY gained.

Discussion

These findings conclusively show a lack of benefit for quality of life or monthly headache days from a brief group education and supportive self-management program for people living with chronic migraine or chronic tension-type headache with episodic migraine.

Background

Process evaluation is increasingly common alongside complex randomised controlled trials (RCTs). This evaluation helps in understanding the mechanisms of impact and how the study processes were executed, and it includes any contextual factors which may have implications for the trial results and any future implementation. This process evaluation is for the Chronic Headache Education and Self-management Study (CHESS) RCT, which is evaluating an education and self-management group behavioural intervention for people with chronic headache. Chronic headache is defined as headaches which are present for ≥15 days per month. The most common types are chronic migraine and chronic tension-type and medication overuse headaches.

Methods

We will use a mixed-methods approach. Quantitative data will be taken from routine trial data which will help us to assess the reach of the study (i.e. did we reach those whom we expected and from where?). Intervention attendance (dose received) and attrition and qualitative data will augment our understanding about reasons why people may not wish to take part in or failed to attend sessions. Interviews with intervention facilitators and trial participants will gain different perspectives on taking part in the trial.

Fidelity will be assessed through listening to audio-recordings for adherence to course content and competence of the facilitation of a sample of sessions.

Discussion

Our process evaluation will allow us to gain insight into how the trial was delivered, the obstacles and enablers encountered and the possible reasons why the interventions may or may not be effective.

Background

The Chronic Headache Education and Self-Management Study (CHESS) multicentre randomised trial evaluated the impact a group education and self-management support intervention with a best usual care plus relaxation control for people living with chronic headache disorders (tension type headaches or chronic migraine, with or without medication overuse headache). Here we report the process evaluation exploring potential explanations for the lack of positive effects from the CHESS intervention.

Methods

The CHESS trial included 736 (380 intervention: 356 control) people across the Midlands and London UK. We used a mixed methods approach. Our extensive process evaluation looked at context, reach, recruitment, dose delivered, dose received, fidelity and experiences of participating in the trial, and included participants and trial staff. We also looked for evidence in our qualitative data to investigate whether the original causal assumptions underpinning the intervention were realised.

Results

The CHESS trial reached out to a large diverse population and recruited a representative sample. Few people with chronic tension type headaches without migraine were identified and recruited. The expected ‘dose‘ of the intervention was delivered to participants and intervention fidelity was high. Attendance (‘dose received‘) fell below expectation, although 261/380 (69%) received at least at least the pre-identified minimum dose. Intervention participants generally enjoyed being in the groups but there was little evidence to support the causal assumptions underpinning the intervention were realised.

Conclusions

From a process evaluation perspective despite our extensive data collection and analysis, we do not have a clear understanding of why the trial outcome was negative as the intervention was delivered as planned. However, the lack of evidence that the intervention causal assumptions brought about the planned behaviour change may provide some insight. Our data suggests only modest changes in managing headache behaviours and some disparity in how participants engaged with components of the intervention within the timeframe of the study. Moving forwards, we need a better understanding of how those who live with chronic headache can be helped to manage this disabling condition more effectively over time.

Background

Patient and public involvement (PPI) plays a crucial role in ensuring research is carried out in conjunction with the people that it will impact on. In this article, we present our experiences and reflections from working collaboratively with patients and public through the lifetime of a National Institute for Health and Care Research (NIHR) programme grant, the Chronic Headache Education and Self-management Study (CHESS), which took place between 2015 and 2020.

Patient and public involvement over the course of CHESS

We worked closely with three leading UK migraine charities and a lay advisory group throughout the programme. We followed NIHR standards and used the Guidance for Reporting Involvement of Patients and the Public checklist. We consulted our PPI contacts using a variety of methods depending on the phase of the study and the nature of the request. This included emails, discussions and face-to-face contact.

PPI members contributed throughout the study in the programme development, in the grant application, ethics documentation and trial oversight, during the feasibility study in supporting the development of a classification interview for chronic headache by participating in a headache classification conference, assessing the relevance and acceptability of patient-reported outcome measures by helping to analyse cognitive interview data, and testing the smartphone application making suggestions on how best to present the summary of data collected for participants. Due to PPI contribution, the content and duration of the study intervention were adapted and a Delphi study with consensus meeting developed a core outcome set for migraine studies.

Conclusions

The involvement of the public and patients in CHESS has allowed us to shape its overall design, intervention development and establish a core outcome set for future migraine studies. We have reflected on many learning points for the future application of PPI.