Discussion
We developed and delivered a novel SMP that was acceptable (to participants and HCPs) in a group and community-based format delivered by those HCPs likely to deliver the intervention in practice. The novelty of our approach included group sessions being spread out further across a longer period than traditional SMPs, with earlier sessions being closer together in time to allow group cohesion to take place.
We did not observe a difference in our primary outcome of health status measured by the CAT. Both groups had an improvement of 1.4 and 1.9 points for the intervention and control groups, respectively (a change of 2 is deemed clinically relevant20). It may be that the intervention was not intense enough to make a difference in this outcome. However, we did demonstrate an improvement in self-reported patient activation (at 6 and 9 months), knowledge (at 6 months), mastery (at 6 and 9 months) and fatigue (at 6 months), in the intervention group compared with usual care. PAM scores improved more than the minimum important difference (MID) of 4 points25 at 6 and 9 months, compared with baseline in the intervention group. This would indicate improved activation and increased self-management skills. There were no statistically significant improvements in anxiety and depression, exercise capacity, dyspnoea or emotion at 6 or 9 months. However, clinically relevant improvements were noted for dyspnoea in both groups (more than the minimal important difference of 0.5 at 6 and 9 months26) and for the ESWT at 6 months in the intervention group this was approaching the MID of 174 s.27 Qualitative results demonstrated that the intervention was acceptable to patients who took part in the intervention and that the SPACE for COPD manual was well received. HCPs also valued the intervention, suggesting it might be best delivered earlier on in the disease process.
MRC and CAT scores demonstrated that overall, participants were milder and less symptomatic compared with a typical PR population.28 The median MRC score was 2, and our median baseline CAT score was 14 (IQR 10–20; mean 14.6). As a comparison, a PR audit conducted around the same time as this study29 noted a median CAT score was 22 (IQR 16–28) and 66% of patients were MRC 3 or 4. These numbers are also comparable to participants in the Kon et al study20 determining the MCID for CAT, where the mean CAT score at baseline was 21.4. As our scores are significantly lower, it is possible that we achieved a floor effect in the intervention group whereby change may be unlikely to be achievable. Other studies have also demonstrated similar improvements in intervention and control groups for QoL measures, for example, Chaplin et al.30 As such, it could be hypothesised that attending a specialist respiratory assessment and being part of a research study could increase the likelihood of patients seeking information about, and take a more active role in, self-management of their COPD (including increased confidence to undertake physical activity or training).14 31 Conducting exercise tests as part of this assessment would have further increased confidence to undertake physical activity or training and as such were not included in the baseline assessment for control group participants.32
Patient activation (defined as patients’ knowledge, skills and self-efficacy regarding self-management33) is shown to play a central role in COPD self-management behaviours and is a central component in the Chronic Illness Care Mode.34 It is important to help patients put knowledge and skills into practice in daily life rather than simply telling them what they need to do35—we went some way to doing this with the format of the group sessions. Goal setting was carried out between sessions that were gradually spaced further apart so that participants were implementing these goals in their daily lives for longer, still with the promise of feedback to HCPs on returning to a group session. However, qualitative feedback suggests that 6 weeks may have been too long and that 4 weeks may have been more suitable. In the intervention group, PAM scores increased on average by almost five points, similar to findings in other self-management studies.36 37 Despite this manuscript describing a lighter touch intervention. This is an important finding as activated patients are more likely to report improved health-related behaviours and clinical outcomes, and partake in collaborative decision-making with HCPs.38 39 Furthermore, each point increase in PAM score correlates to a 2% decrease in hospitalisation and 2% increase in medication adherence.40 Therefore, these data indicate that the intervention has the potential to reduce healthcare use.
Disease-specific knowledge increased postintervention supporting the effect of the intervention and at 3 months post intervention, this knowledge was retained (slight increase in scores between 6 and 9 months in both groups). Reduction of disease-specific knowledge in the months following the completion of SMP and rehabilitation programmes is a common finding,41 therefore, we were encouraged to see this outcome maintained at 9 months. New international guidelines could offer some guidance when designing the education component of self-management and rehabilitation programmes in the future.42
Strengths and limitations
Offering a community-based programme, at venues and times more suitable for participants, went some way to remove common barriers to attendance. A high number of intervention participants could be classed as completers of the intervention (mean number of group sessions attended was 3.88 and 62.5% completed four or more sessions)—this, alongside qualitative findings, seems to indicate that conducting sessions in the community, local to the participant, at times suitable to them, better enabled participants to engage with and complete the SMP. It is hard to assess if the adherence rates to the intervention in this study are comparable to that of one-to-one sessions in Mitchell et al14 as this was not measured. However, both studies had high completion for the 6 month assessments from participants in the intervention groups (Mitchell et al14 80%, this study 83%). Our inclusion criteria allowed for recruitment of a wide range of participants, not just those who would usually be suitable for PR. Indeed, the median MRC of our recruited population was grade 2. This group is often excluded from PR programmes.28 The intervention was underpinned by social cognitive theory43 and motivational interviewing techniques.44 It also included other techniques such as goal setting and solution-focused goal feedback that have been shown to be effective in modifying behaviour. Further, it was delivered at an intensity that was possible to incorporate into routine practice. The way in which the group sessions were delivered maximised support from/contact with an HCP and allowed for rapport to develop between participants and HCPs in the initial stages of the intervention, prior to longer breaks between sessions.
For some intervention participants, distance from group venues was a problem, causing withdrawal from the study. This is because they may have had to attend sessions that were outside of their local area as we were recruiting in multiple areas at the same time. For the same reason, other participants were waiting up to a month for group sessions to start. Future studies may wish to consider cluster randomisation over individual participant randomisation to avoid these issues. In practice, this would occur naturally as GP would most likely run SMPs for their patients, ensuring a venue that is suitable for all. While group-based interventions have some advantages in terms of peer-support and likely cost savings; some patients dislike group formats. Also, this study was conducted prior to the start of the COVID-19 pandemic which limited in-person group-based interventions. Future implementation work therefore may need to consider these factors. It is likely that a choice of formats (eg, one to one, group-based, in-person, remote) would help to improve uptake16 and engagement from both patients receiving the intervention and services providing this.
Results suggest that the CAT may not have been the most suitable primary outcome measure for patients with milder disease.8 It may be that a general self-management measure (eg, HEI-Q45) or the PAM are more suitable. These would also measure the important behaviour change aspects of the SMP.46 While we have not compared the outcomes from this study to those delivering the SPACE for COPD intervention on a one to one basis; we were able to show comparable changes in the CRQ-dyspnoea domain (the primary outcome in the Mitchell et al study14) at 6 months in the intervention groups: 0.5 in the current study and 0.66 when delivered 1:1, both meeting the MCID. The Mitchell et al study did not measure the CAT.
Qualitative results indicate acceptability of the intervention on behalf of the patients’ taking part and the HCPs delivering the sessions. Although the focus groups were only completed in those available at the end of the intervention period, all patients were invited to take part regardless of the number of sessions they attended in total. Despite the qualitative results suggesting that the intervention was acceptable, there were a large number of participants who did not respond to the study invitation (81.5%) and a further 36% who declined to take part. Unfortunately, a number of participants were also withdrawn by the study team that fell within the intervention arm of the study. Although the study was powered for the CAT at 6 months, recruitment was uneven between groups. Reasons for withdrawal were not related to the intervention in the vast majority of cases—many were ineligible (did not have a diagnosis of COPD—as a consequence, the sample size was reassessed and additional participants recruited.) For those who were eligible, social problems and other health problems were often cited as reasons for withdrawal. In many cases participants withdrew from the study before commencing the intervention.
Future implications
The importance of self-management is widely acknowledged in people living with COPD3–5 and opportunities should be maximised from the time of diagnosis through to more severe disease.47 However, there is currently no provision for a structured supported SMP in the UK for those with milder disease. Although mild disease was not an eligibility criterion for this study, we did primarily recruit those in GOLD stages I and II with a median MRC grade if 2. SPACE for COPD as a group-based community situated intervention could therefore be offered to newly diagnosed patients with COPD, with less severe problems and those with capacity to manage their disease, to bridge the gap between diagnosis and offering of PR. For those patients with more severe disease, and may have comorbidities, PR would assume greater importance.47 This prevention model has only recently been adopted in COPD, with calls for interventions to reduce risk in people with early disease48 and potential for considerable health and health service gains if we could facilitate self-management support in patients with early disease and slow their decline.49 In order to establish whether this is possible, long-term follow-up studies are required.
Opportunities to improve self-management skills should also be embedded in PR programmes. In the future, there may be an opportunity to explore the value of the SPACE for COPD programme alongside rehabilitation, or indeed, an alternative for those unwilling or unable to attend. Currently the programme is also being tested as a group based intervention for those who have recently completed PR as a maintenance tool.50 The SPACE for COPD group-based SMP offers patients a choice which may be more convenient for those who may still be working and unable to commit to traditional PR, or for those struggling to travel and are only able to travel in the local area. The nature of the model also allows for patients to access specialist teams for a longer period of time but also encourages the group to meet outside of the scheduled sessions, creating in effect a local COPD community with shared interests. In future, a more intense version of the programme may be warranted to increase gains in the outcomes of interest.
