GLA:D^® Back: implementation of group-based patient education integrated with exercises to support self-management of back pain - protocol for a hybrid effectiveness-implementation study

This clinical intervention is intended for people with persistent or recurrent low back pain (LBP) and in need of improved self-management. The content and intervention mapping are presented in detail elsewhere [30]. Briefly, GLA:D Back consists of an individual session of clinical testing and goal-setting at the beginning and end of the program, two one-hour group sessions of patient education, and 8 weeks of twice-weekly one-hour supervised exercises sessions (Fig. 2).

The overall aim of GLA:D Back is to improve the participant’s ability to self-manage and the content is based on a cognitive behavioural approach aimed at supporting pain self-efficacy. Theoretically, the proposed effect mechanisms are that the knowledge and skills achieved during participation in the program translate into changed believes and performance, which in turn will improve self-efficacy, decrease disability, increase quality of life, and ultimately reduce health care utilisation and sick leave (Fig. 3).

Clinicians can adapt GLA:D Back to their particular setting and to the needs of individual patients, however, the following core parts cannot be altered: 1) an individual session at the beginning and the end of the program with goal-setting and physical tests, 2) 2 h of patient education and 16 sessions of supervised exercises over 8 to 10 weeks, 3) key messages stating that pain is not a sign of danger, 4) an explanation of back pain using a behavioural model of (im) balance between demands and capacity rather than emphasising tissue damage (Fig. 4), 5) guidance for patients in exploring movement rather than being taught to perform exercises in one ‘correct’ manner, and 6) entry of data into the clinical registry. Refinements of the content of the patient education or exercise program will be made continuously by the research group without violating these principles.

Clinicians are trained in a two-day course (7 h each day) delivered by the developers of the program (AK, IR, PK, JH) and by employed expert clinicians. The course program can be seen in Additional file 1.

The primary aim of the course is to motivate clinicians to adapt a behavioural rather than a biomedical/structural treatment orientation and to provide tools that support the implementation of the GLA:D Back program. The GLA:D Back course was developed using the COM-B model in order to supports clinicians’ behaviour change by improving their capability, motivation and opportunity to change. The learning goals are pursued by using different educational elements including education through lecturing, persuasion and modelling by examples, in addition to skills training (Table 1).

First, participants acquire an understanding of the content of the clinical intervention and the arguments for standardised care. They are presented with an overview of current evidence about the burden of LBP, evidence supporting persistent LBP as a non-injury pain condition, a brief introduction to pain modulation and the cognitive behavioural model [39], and a summary of the evidence base for GLA:D Back. Clinicians are informed that GLA:D Back was developed for people with persistent or recurrent non-specific LBP who have a need for improved self-management, and the decision as to whether or not a patient enters the program is left up to their clinical judgement and the patient’s motivation.

Second, the main body of the course is aimed at developing participants’ ability to deliver the GLA:D Back clinical intervention as a way of addressing pain cognitions and behaviours by introducing all elements of the clinical intervention: goal-setting, clinical tests, patient education and supervised exercises. Goal-setting is introduced by presenting the SMART-model (Specific, Measurable, Acceptable, Realistic, and Time bound goals) [40]. Role plays to practice the use of explanatory models and skills training are used to become familiar with pain education, physical tests and exercises. Using some of the slides from the patient education material, participants work in groups taking turns to deliver key messages from the pain education to each other as they would in a real patient session, evaluated using peer feedback. The clinical tests and exercises are introduced in a practical session where tests are carried out on a colleague, and exercises are performed by the participants while examples of variations are presented. The supervision of exercises is practised in a session based on video cases and centred on the messages (spoken and unspoken) that can be delivered while supervising patients in the clinic. Key messages from the patient education should be reinforced during the instruction of exercises, for example, when explaining to patients how to cope with pain provocation during exercises rather than withdrawing from the activity. Finally, participants are introduced to the clinical registry used for data entry. At the end of Day 2, clinicians plan how and when to start GLA:D Back in their clinics. They are encouraged to pursue implementation by referring to the success of the GLA:D Knee/Hip model and the experience with GLA:D Back from pilot testing, as well as by emphasising the importance of physiotherapists and chiropractors taking on the role of ensuring access to evidence-based care for back pain.

Courses may be refined during the process of implementation, however, the core parts that are not subjects to change are: 1) familiarising participants with the key messages of the clinical intervention, 2) the introduction to the clinical registry, and 3) that the clinicians practice the delivery of patient education and the instruction of exercises during the course.

The national implementation will be driven by clinicians who have attended the course and are trained in GLA:D Back. Therefore, the first step was to generate awareness of the development of GLA:D Back in professional journals, at seminars and meetings, and by mentioning the new initiative at the GLA:D Knee/Hip courses. Then, in August 2017, nine clinics (31 clinicians) participated in piloting the GLA:D Back program and took their own initiatives to promote their participation via websites, social media and contact with GPs.

In February 2018, when course registration opened for all interested chiropractors and physiotherapists, the Danish Physiotherapy Association and the Danish Chiropractors’ Associations and some physiotherapy special interest groups were asked to advertise the program to their members. The GLA:D Back website, and an invitation to register, were also promoted by the research team on Facebook, LinkedIn and Twitter and to clinicians who use the GLA:D register for knee/hip patients when they logged on to the registry (Table 2). Course participation is at a cost of approximately €550 (2018).

Clinicians who participate in the course and decide to deliver GLA:D Back at their clinic are listed as certified clinicians on the GLA:D Back website (http://​gladryg.​sdu.​dk/​). Because GLA:D is a registered trademark, only clinicians trained at SDU can use the brand, and sustained certification comes with the requirement to enter all appropriate patient information into the clinical registry. Updates and re-certification are planned for the future.

To support the implementation after the course, clinicians are given access to detailed patient educational materials, exercise programs, and written suggestions of language that can be used in the supervision of exercises after the course. They also are given GLA:D Back logo t-shirts and posters for the clinic with key messages from the patient education and an overview of the exercises [30]. To assist a uniform promotion of the program, clinicians get information leaflets directed at patients and at GPs that can be used with their individual clinic names and logos.

There are no strict patient inclusion criteria, and the decision to enrol a patient in GLA:D Back is at the discretion of the clinician in a dialogue with the patient, when clinicians judge that the patient would benefit from improved self-management skills. Patients can be enrolled at the first visit or later depending on whether individual treatment sessions are deemed necessary prior to starting GLA:D Back. When patients begin the program their contact information is entered in the clinical registry, which in turn activates links to patient questionnaires that are sent automatically on the day of registration and again after 3, 6 and 12 months. Clinicians enter results of physical tests and patients’ individual goals in the registry before patients start the GLA:D Back intervention and when the program ends. The price and out-of-pocket expenses for the GLA:D Back intervention can vary between clinics and types of insurance.

Consultants for physiotherapy, chiropractic, and family medicine working with primary care in the five administrative regions were invited to an information meeting about GLA:D Back before course registration was opened (January 2018). Family physicians will be informed through professional journals and seminars.

Data sources will include: 1) data obtained from clinicians when they register for a GLA:D Back course, 2) clinician-completed surveys before and directly after the course and after 5- months, 3) observations of patient education and exercise group sessions in selected clinics, 4) patient physical tests at baseline and at the end of treatment, 5) patient-completed surveys at baseline, 3-months, 6-months and 12-months, 6) patient interviews, and 7) information from Danish registries (Fig. 5). Patient and clinician data are collected electronically using the REDCap software (Vanderbilt University) provided and supported by the Odense Patient data Explorative Network (OPEN) [41].

The clinical intervention will be evaluated based on data about the patients who enrol in the GLA:D Back program, i.e. those patients who are registered in the GLA:D Back registry.

Due to the lack of a control group we will compare mean changes in these patients with those observed in the intervention groups of trials that have demonstrated positive effects of patient education and exercise therapy. We will look for trials with little risk of bias that included patient populations comparable with those enrolled in GLA:D Back.

Health care utilisation and sick leave will be investigated using national registry data to compare individual patterns of health care utilisation (primary care visits, hospital visits due to LBP, imaging for LBP, prescriptions for pain medication) and sick leave from 1 year before participation in GLA:D Back to 1 year after participation (Fig. 5).

Clinicians’ beliefs about back pain and confidence in managing patients with LBP (described below) and their baseline characteristics will be investigated as potential predictors of patient outcomes. Potential mechanisms of change in patients will be investigated by testing the mediating pathways outlined in the theoretical model (Fig. 3). We will specifically test the extent to which improved illness beliefs, increased perceived ability to perform exercises, and improvements on physical tests may explain a reduction in fear avoidance, improvement in expectations and increased self-assessed physical capacity during the 3 months after inclusion and during the following 3 months (3- to 6-months follow up). We will then test to what extent reduced fear avoidance beliefs, improved expectations and increased physical capacity during the initial 3 months relate to an increase in self-efficacy during the same period and to improved self-efficacy after 6 and 12 months. Next, we will investigate the relationship between increased self-efficacy and outcomes of self-management success, and whether self-efficacy is a mediator between beliefs and performance and success in self-management. Lastly, we will test to what extent improvements in disability, quality of life and pain after 6 months predict health care utilisation and sick leave after 12 months, to what extent improved self-efficacy influences these outcomes, and if this is fully or partly explained by improvements in self-management.

Patients’ perceptions related to developing pain self-efficacy and ability to self-manage will be explored using individual interviews that address back beliefs, perceived ability to manage LBP, and what aspects of the LBP experience need to change during and after participation for patients to consider the intervention beneficial.

All clinicians participating in a course will be included in this evaluation. The course content and delivery will be evaluated via observations by an expert in medical education to confirm that the planned learning activities are delivered. Outcomes of learning activities (Table 1) will be compared between courses and any trends in change over time described.

Evaluations of clinician-level outcomes will be investigated in an observational longitudinal design with clinician-reported data collected before the clinician training course, immediately after the course and 5 months later, and with patient-reported information on delivery collected at their three-months follow up. Changes over time in treatment orientation and confidence in managing patients with LBP will be described, and potential relationships between clinician characteristics, treatment orientation and confidence will be determined.

The implementation process will be evaluated, starting with the promotion of courses for clinicians, the degree of course participation, the adoption and fidelity of the implementation in the clinics, through to the potential impact on health care utilisation at a national level. An evaluation of sustained implementation is not part of this protocol. Elements of the process evaluation are listed in Table 3.

Promotion of courses and pre-defined minimum standards for each element of the promotion strategy will be monitored during the roll-out (Table 2). A research team, not otherwise involved in GLA:D Back, will conduct the monitoring and provides feedback to the GLA:D Back project group at regular intervals, which in turn will determine the actions required when standards are not met. For example, if one region has less than 60 clinicians registered, this will be fed back to the project group with information on any particular clinician groups (e.g. related to professional background or clinical experience) that appears not to have been reached so that actions can then be taken to reach that specific group(s).

The adoption of GLA:D Back will be investigated by measuring the extent to which clinicians who have participated in the GLA:D Back course conduct the program in their clinics. Individual clinician factors and organisational factors will be investigated as determinants of adoption.

Fidelity will be investigated quantitatively by asking about treatment content and recall of key messages via patient questionnaires, and by ensuring that patients’ individual goals are registered. The quality of care in terms of delivering the key elements of the intervention will be further explored by video-based observations and from patient interviews that will provide a patient perspective on the content of care.

The profile of patients enrolled in the GLA:D Back programme will be compared between clinics and administrative regions to investigate if clinicians seem to agree on the target group as well as detect if regional politics or planning affects patient selection. Also, the proportion and profile of patients enrolled who do not complete the program will be compared across clinician profiles and regions. Finally, we will ask clinicians about their experience with patients being motivated to enrol in the program and the reasons why patients considered suitable candidates for the program declined participation.

The frequency of referrals to imaging and to secondary care for back pain, opioid prescription and long-term sick leave related to LBP at the national level will be compared between 2016 and 17 before implementation starts and during 2020–21, i.e. commencing one year after all Danish regions have had clinicians educated. Comparisons will be made at the overall population level as well as for populations with a history of LBP identified from sources such as The Danish National Health Survey.

An overview of the study measures is provided in Table 4.

An overview of the study measures is provided in Table 5.

Information will be collected to describe clinician profiles: age, sex, profession, years of clinical experience, type of employment, experience with GLA:D knee and hip, and whether or not their clinical work is conducted under a contract with public health insurance. Information will also be collected about clinics: the number of physiotherapists, chiropractors, massage therapists and other clinicians in the clinic, whether GLA:D for knee and hip is offered, and if back pain group exercises were offered before introducing GLA:D Back.