Statistics from Altmetric.com
Strengths and limitations of this study
The approach enabled us to consider and integrate fidelity assessment.
We were able to modify recruitment processes and the course to accommodate patient needs.
We used the feasibility study to train the team and engage new networks for the main trial
The approach was resource intensive and lengthy but contributed to an efficient trial.
Chronic musculoskeletal conditions are costly and burdensome to individuals and the society.1 Point estimates of the prevalence of chronic musculoskeletal pain range from 46% to 76%.2 Despite an increased understanding of the factors contributing to the development of chronic pain, there has been little improvement in how successfully it is treated and managed.3 Treatment centres around pharmaceutical agents and physiotherapy. More complex interventions such as pain management programmes delivered by multidisciplinary teams and self-management courses delivered by lay people with chronic pain are also used to address the complexity of living with and managing chronic pain.4–6 The UK Department of Health and The Health Foundation have invested in the implementation of lay-led (ie, peer-led) self-management training courses through the Expert Patients Programme (EPP)6 ,7 and the cocreating health initiative that aims to help people help themselves.8 The available evidence, however, suggests that it may not reduce healthcare resource use as expected9–11 and that there are only modest short-term beneficial effects on other outcomes. Very few studies have examined the long-term effects.10 ,11
In response to the paradox of continued government support of self-management programmes and equivocal evidence of effectiveness, the COping with persistent Pain, Effectiveness Research into Self-management (COPERS) study was commissioned by the UK National Institute of Health Research as a 5-year programme grant to improve the self-management of chronic pain.
Our aim was to design and test a practical and acceptable self-management intervention for chronic musculoskeletal pain.
This study illustrates how the Medical Research Council (MRC) framework for developing complex intervention can be implemented and used to develop interventions.12 We used the recommended approach for developing and designing this new intervention, which consisted of three phases
Identifying the evidence base,
Identifying appropriate theory to inform and model the design of the intervention,
Feasibility testing the intervention.
The first two projects informed the design of a pain self-management course, which we then pilot tested.
Phase I: Identifying the evidence base
We conducted two systematic reviews (SRs) to identify effective components and characteristics of pain management courses (SR1) and predictors, mediators and moderators of outcome in pain management courses (SR2). The methods and results are presented in detail elsewhere.13 ,14 Additionally, we did a qualitative study (QS) of people living with chronic pain.
We searched relevant databases including: MEDLINE, CINAHL, AHMED and PsychInfo from January 1994 to April 2009 for randomised controlled trials (RCTs) and SRs of self-management interventions. We defined self-management programmes as structured, taught or self-taught courses with distinct components principally aimed at patients (rather than carers) with the goal of improving the participants’ health status or quality of life by teaching them skills to apply to everyday situations. To be considered a ‘programme’ (which implies more than one component), the intervention had to contain at least two of the following components: psychological components (such as behavioural or cognitive therapy), mind-body therapies (such as relaxation, meditation or guided imagery), physical activity (any form of exercise), material on lifestyle (such as dietary advice and sleep management) and pain education (such as understanding their condition and how to take medication effectively).
We characterised the interventions according to: type of delivery (group, individual, mixed or remote (eg, web based)), type of tutor (healthcare professional, lay or a combination of tutors), setting (medical (ie, hospital, physician office or primary care), community or work based), duration (more or less than 8 weeks) and number of different components.
We examined the following outcomes: pain intensity, physical function, general mental health, depression, anxiety, social function, healthcare use, global health measures, quality of life and self-efficacy, but only examined outcome measures with published evidence of validity and reliability. We grouped outcomes into three follow-up intervals: short term (<4 months), medium term (4–8 months) and long term (>8 months).
We used random effects model meta-analysis to generate standardised mean differences and grouped data according to the presence or absence of course characteristics or components. We looked for patterns of clinically important and statistically significant differences between groupings across different outcomes and follow-up intervals.
Effective components and characteristics of pain management courses (SR1)
Overall, the literature indicated that the strongest evidence for pain outcomes was for group self-management programmes led by healthcare professionals while lay-led courses appeared to benefit participants’ self-efficacy (see table 1).13
The duration of courses did not appear to significantly influence their reported effectiveness and the setting had little impact on the outcome. Interventions including each of the components we tested (except mind body components) showed beneficial effects. Psychological components showed the most benefit. Increasing the number of components did not improve the overall benefit. Overall, our analysis provided some evidence for the use of short, group-delivered courses in convenient settings that have healthcare professional input (more detailed information can be accessed via reference 13).
Predictors, mediators and moderators of outcome in pain management courses (SR2)14
We identified papers that included analyses of predictors, moderators or mediators. We did meta-regression analyses using the standardised mean difference and estimated the precision of associations.15
We found strong evidence that self-efficacy and depression at baseline predicted outcome and strong evidence that self-efficacy and potentially pain-catastrophising and physical activity mediated outcome from self-management programmes for chronic musculoskeletal pain. There were no data on moderators of treatment.14
We conducted a phenomenological qualitative interview and focus group study. For the interview study, we purposively sampled course participants by gender, age and high-course and low-course attendance from two EEP providers and one pain management course provider, and for the focus groups we recruited a convenience sample of experts from those who had recently published in this field and tutors from two local EPP providers. We carried out in-depth interviews with 16 chronic pain self-management course participants (11 were female, 7 were 45 years or over, 10 were white, 6 were of South Asian origin and 10 people had attended half or more of a course). We convened two focus groups, one with self-management ‘experts’ (n=5) and another with course tutors (n=5). Topics discussed included referrals attendance, course content, course delivery issues and expectations. We used a thematic framework approach to analyse the data and identified key themes and subthemes relevant to devising a new pain self-management course. Participants identified six key areas that they felt their treatment for their chronic pain should address; it describes the participant needs and expectations (table 2). We tried to incorporate these needs and expectations as much as possible into the course. While we could not achieve some things by running a course, we aimed to give participants the skills to realistically assess their needs and show them how to achieve them.
We noted that the patients with chronic musculoskeletal pain whom we interviewed made positive comments about the courses they attended and had particular traits. They were prepared to be socially engaged; they were motivated prior to the course and took up new activities (not necessarily exercise related) after the course. Participants liked the social element and the relaxation components of the course. Good course facilitation and social support cemented their experience. Those with a low mood, poor social skills and unwillingness to change/reflect seemed less likely to engage with these types of courses.
Phase II: Identifying appropriate theory to inform and model the design of the intervention
We searched the literature and spoke to key experts on behavioural change theory and models of persisting pain. We considered the following psychological theoretical models and learning and behaviour modification techniques: social cognitive16 ,17 and cognitive behavioural theory,18 ,19 including psychological flexibility (acceptance and commitment therapy, ie, the acceptance of internal experiences or things that cannot be changed countered by behavioural change techniques that are designed to reorientate people towards meaningful activity,20 ,21 theory of planned behaviour and reasoned action22–24 (including emotional rationalisation) and health belief models. In addition, we looked at attention control techniques25 and physical movement to underpin and inform our intervention. Figure 1 illustrates the relationship between theory and course design.
We established a patient working group to devise an acceptable and appropriate intervention that reflected the evidence we had obtained and the theories we had identified as appropriate.
As recommended by the MRC guidelines, we considered patient pathways through the self-management programme, as well as the likely action and interaction of the different components on outcomes in an attempt to model the impact and effect of our intervention.12 We convened a ‘consensus’ group of two patients, three clinicians (general practitioner (GP) and two clinical psychologists), four researchers and one commissioner to agree on the outcomes we wanted the intervention to affect and the appropriate measurement tools. They were improved function despite pain, ‘better’ healthcare resource use, reactivation into society, more self-confidence in managing pain, better coping and reduced anxiety and depression. The measurement tools were selected and tested in the feasibility study (a sample of which is shown in table 3). We assessed the behavioural change theories relevant to a new intervention and identified individual behavioural change techniques for different learning outcomes and the different components of a course (table 4). We also used a taxonomy of behavioural change techniques developed by Abraham and Michie26 to describe the techniques we adopted to promote positive behavioural change in self-management groups.
The courses also allowed opportunities for people to learn and try new behaviours in an appropriate environment.
Table 4 shows our rationale for mapping and modelling theory to behavioural change techniques and the methods used by facilitators throughout the courses. The final column describes the behavioural change techniques that were used throughout the courses. While some sessions required facilitators to employ techniques focusing on providing feedback, other sessions provided instruction to promote behavioural change and yet others allowed participants to try out techniques within the ‘safety’ of the learning environment and the group. The techniques employed by facilitators were often dependent on the needs of the participants and the groups and therefore are utilised as necessary and when required in each individual session. No negative or coercive behavioural change techniques were recommended or used as part of the courses. We decided that the components of the course should include: psychological concepts using cognitive behavioural (http://www.babcp.com/) approaches to managing chronic pain (these covered: acceptance, attention control, goal setting and action planning, recognising unhelpful thinking and behaviours); the course also covered communication skills, relationships, promoting better sleep, medical education, social networking, hobbies and activities, posture and movement, breathing, relaxation and visualisation and guided imagery. Our QS indicated that patient understanding about pain was limited, so we therefore decided to include an educational DVD with a pain consultant answering common questions from patients with chronic pain. The structure of the course we piloted is shown in table 5.
The findings from the SRs and the QS informed the design of the new intervention (table 1). The final course structure and content included a mix of theoretical concepts and psychological, behavioural, educational and physical techniques. Our qualitative research and evidence reviews identified a variety of components, characteristics and functions to accommodate in the self-management course, hence the eclectic design.
We designed a group course to be facilitated by a healthcare professional (a psychologist, physiotherapist, chiropractor, osteopath, occupational therapist or GP) and a lay person with chronic pain with prior experience in small group facilitation (eg, a course facilitator on the EPP). The course was structured to be delivered over three short days in 1 week (10:00–14:45), with a 2 h follow-up session 2 weeks later. We also designed a 2-day training programme for all potential facilitators. All courses were to be held in a convenient, accessible location for study participants.
Phase III: assessing the feasibility of the intervention (COPERS trial ISRCTN 24426731)
We used an uncontrolled pilot study approach to test the feasibility of delivering the intervention and the receipt of the intervention.
Adults (18 years and over) included those with persistent musculoskeletal pain (pain in their muscles or joints lasting longer than the normal expected healing time of 3 months) who were physically and mentally able to attend a community-based group course. We excluded those who had any other more serious comorbidity than their pain (such as terminal illness, cancer, uncontrolled addictions or other mental health issues).
All patients were required to be fluent in English as this was a group-based course reliant on discussion and interaction.
We estimated that six courses of 8–10 people would be sufficient to evaluate the course; we aimed to recruit around 60–80 participants allowing for dropouts.
Inner London urban community in East London. We recruited patients from two local general practices, the Tower Hamlets Persistent Pain Service and the musculoskeletal physiotherapy service at Mile End Hospital. Participants were identified by clinicians from known regular patients and ad hoc from face-to-face consultations. We delivered the courses in community-based venues convenient for participants.
We used a number of outcome measures to test completion rates and acceptability. Questionnaire instruments are shown in table 3.
Participants completed postal questionnaires at a baseline and at 3 months.
Fidelity and facilitator training: adherence and competence
We trained 15 facilitators to deliver the course. They underwent a 2-day training course and included six lay people with chronic pain and nine healthcare professionals (one chiropractor, three osteopaths, two physiotherapists, one psychologist, one occupational therapist and one GP). Three of these trained healthcare professionals were unable to facilitate courses due to work pressures.
We observed all the courses and wrote extensive field notes. In addition to collecting our outcome data, we asked for written feedback about the courses from the facilitators and the participants and also interviewed 13 participants. We collected data about attendance for every component of the course and reasons for non-attendance and non-participation.
We sent 518 invitation letters to people with chronic pain (335 were female); 68 people agreed to participate (13% response rate), 8 withdrew without accepting a place, and 60 accepted a place on a course (33 were female). In 4 months, we delivered six courses with an average of seven participants per course. Overall, participants attended 85% of all sessions.
The overall satisfaction score (0–5 with five indicating most satisfied) was 4.2. The interview data revealed that participants most liked meeting other people with chronic pain and the relaxation sessions. Participants who attended few sessions (less than 1 day), or did not attend, reported that work commitments and being in a group were an issue, and that some poor facilitation affected their learning and pleasure. Some participants mentioned that they thought the intimacy session was too personal and they did not want to disclose this type of information to people they did not know well.
This centred on the need for preparation and experience in facilitation and handling difficult situations. Observer feedback identified that there were some modules that needed to be amended, and that the intimacy session did not work well due to the reluctance of participants to discuss issues. In addition, the facilitators felt that the ‘buddying-up’ system (the facilitators created and provided the opportunity and time for group members to exchange contact details) should be instigated and initiated by the group themselves, not the facilitators, as some group members found the concept of buddying too intrusive. More handouts were suggested and facilitators needed more training in facilitation to make sure that they were confident. However, all facilitators felt that they improved as they progressed through the courses.
No serious adverse events occurred as a result of the intervention.
Descriptive results from the baseline questionnaire showed that 53% (23/43) of participants were female, modal age range was 41–50 years (15 participants). Twenty-one people (49%) regarded themselves as unemployed or as unable to work due to pain, 12 were employed, 5 were retired and 4 were looking after family at home.
Overall, the profile of the participants at baseline suggested that they had above average pain (6.7 on a scale of 0–10) and rated their quality of life as low; also, they had high anxiety scores and were not coping well (table 3).
The follow-up response rate was 58%. Participant, facilitator and researcher feedback suggested that the follow-up questionnaire length (94 questions) was too burdensome and therefore unacceptable for the main trial.
Findings and recommendations to optimise the intervention and a trial protocol
The course was feasible, acceptable to participants and deliverable. Participants were positive about the course and the content appeared to be meaningful to them. Attrition was very low over the three main days: participants attended an average 85% of the course. Attrition has been reported as an issue in other trials; one such trial (intervention arm n=313) using expert patient programmes over a 6-week period showed a loss of 26% participants between referral and course attendance and 40% of participants attended three or less sessions overall, that is, 50% or less of the course sessions.11
The facilitation and group process may have optimised the learning process as discussion embedded participant thinking. All the course evaluation material suggested that good facilitation skills were crucial for positive participant perception. Comprehensive facilitator training is essential for courses to run effectively.
Recruitment to the study was difficult and conversion rates from invitation to course attendance were lower than we had hoped (∼13% of those invited), but were in line with other studies of this nature recruiting patients from primary care with chronic conditions.32–34 Despite this, we had sufficient interest from patients to run six courses and this feasibility study showed that there was a demand for learning about non-pharmacological approaches to managing pain. Procedures for future recruitment can be enhanced by increasing the number of invitations and devising and testing a comprehensive and inclusive electronic search strategy for patients with chronic pain.35 It is also quite likely that recruitment to the intervention would be higher outside the context of a research study.
Recruitment needs to be timed with the delivery of the intervention (ie, course dates which had to be planned in advance due to facilitator and accommodation availability). We estimated that an average lead time from identification of participant, screening for suitability by GP, sending invitation, receiving enquiries and interest, sending out and returning baseline questionnaires, randomisation and booking on a course takes around 8 weeks. Building rapport with participants from the outset is crucial to reduce loss of potential participants prior to being enroled, randomised and/or booked on a course.
In this pilot, participants reported poor quality of life, low self-efficacy to manage their chronic pain, relatively high levels of social isolation, poor coping and a tendency to anxiety and depression. Thus, secondary outcome measures need to reflect these health and social states. Our descriptive baseline data may explain why the ‘buddying system’ and the ‘Intimacy session’ from the course were too difficult for some participants to deal with. It may be that depression should be addressed with patients prior to, or in conjunction with, attending these types of courses.
The theory underpinning the decisions to include the variety of sessions and behavioural change techniques worked within the group learning environment; this has also been shown to be effective in other studies of chronic pain.33 ,36 The learning sequence we adopted enabled each session to build on the previous session, and in many cases the participants were able to predict the next phase of learning in advance. The learning and flow of information was pitched at a level where participants could follow the structure and understand the content. This was shown in the daily feedback sheets where we asked participants what they had learnt; their learning mapped well onto the learning objectives. The quotes below illustrate some of this.
it was shocking to think there is no cure for my pain but I suppose there isn't, otherwise I would have been given it by now
the ‘unhelpful thinking’…they were all me, I will look out for these now, I've stuck them (the list of automatic thoughts) on my fridge door
the relaxation and the breathing really help me
I spend all day trying not to think about my pain but that's the worst thing I can do
The best part was meeting everyone
The participants valued the social interaction on the course highly; for some participants, it appeared to have an impact on self-esteem and confidence, for example, self comparison with others in less fortunate circumstances, perspective on life, distraction, laughter and release from boredom and isolation in some cases.
We found that it was possible to train both lay and non-psychologists to facilitate the courses and deliver cognitive behaviour approach. Delivery styles did vary and there is value in thorough training and evaluation of training and subsequent delivery of courses, embedding fidelity assessment from the outset to measure the adherence and competence of those delivering the intervention. We found that the course stood up to the inexperience of our facilitators to deliver an entirely new course; the content in terms of the discussions, information and handouts was robust enough to make an impression regardless of the delivery style. We recommend that inexperienced personnel are partnered with experienced personnel initially.
A by-product of testing the programme was staff training and development based on their experience of conducting the pilot. We also found that we built valuable networks and contacts, which was helpful for the main trial.
The MRC guidance for developing complex interventions enabled us to develop and test an evidence-based and theory-informed pain self-management course. The process enhanced the intervention and gave the study team confidence in the modified intervention and trial procedures and processes necessary to run a full effectiveness and cost effectiveness RCT efficiently.37
Contributors DC wrote the first draft of the paper and managed all the development phases of the study. KH worked on all phases of the development and has contributed to and commented on the manuscript. MU and SJCT are the principal investigators on the project and provided oversight and guidance in all stages of the research and contributed to the manuscript. TP was involved in the systematic reviews and the theoretical development of the programme and has commented on the paper. AR developed and delivered the education component of the course and has commented on this paper.
Funding This paper presents independent research funded by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research scheme (RP-PG-0707–10189). The views expressed in this publication are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.
Competing interests None.
Ethics approval Ethics approval was granted by the East London and The City Research Ethics Committee Alpha (10//H0704/7). ISRCTN: 24426731 COPERS trial.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement All additional data pertaining to this study will be published in a full NIHR report published by the funder.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.