Increasing uptake of self-management education programmes for type 2 diabetes in primary care: the Embedding research programme including an RCT
Background Self-management education and support programmes help people with type 2 diabetes to manage their diabetes better. However, most people do not attend these programmes. Objective Increase type 2 diabetes self-management programme attendance. Design Workstream 1: develop intervention (mixed...
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
NIHR Journals Library
2025-02-01
|
| Series: | Programme Grants for Applied Research |
| Subjects: | |
| Online Access: | https://doi.org/10.3310/KWYF5914 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Background Self-management education and support programmes help people with type 2 diabetes to manage their diabetes better. However, most people do not attend these programmes. Objective Increase type 2 diabetes self-management programme attendance. Design Workstream 1: develop intervention (mixed methods). Workstream 2: refine intervention and trial design (feasibility study). Workstream 3: evaluate effectiveness (18-month wait-list cluster randomised controlled trial with ethnography component; baseline: months −3 to 0; step one: months 1–9; step two: months 10–18; minimum clinically significant difference in glycated haemoglobin: 1.1 mmol/mol; target sample size: 66 practices). Workstream 4: health economics analysis; 12-month observational follow-up of trial population; qualitative substudy. Setting Primary care practices and providers of self-management programmes (East Midlands, Thames Valley and South Midlands, Yorkshire and Humber). Participants Workstream 1: 103 stakeholders. Workstream 2: 6 practices. Workstreams 3–4: 64 practices (92,977 people with type 2 diabetes). Qualitative substudy: 30 participants. Intervention Embedding Package (marketing strategy for self-management programmes; user-friendly referral pathway; new/amended professional roles; resources toolkit) delivered through an online portal for practices and providers (‘toolkit’; 88 live accounts; average of 19 page views/week); people working with practices and providers to embed self-management programmes into routine practice (‘embedders’). Additionally, a patient digital support programme (MyDESMOND) was developed. The comparator was usual care. Main outcome measures Patient-level glycated haemoglobin (primary outcome, continuous, mmol/mol) and referrals to, and attendance at, self-management programmes (main secondary outcomes; binary yes/no variables) compared between control (wait-list: baseline and step one; immediate: baseline) and intervention (wait-list: step two; immediate: steps one and two) conditions. Data sources Existing interviews, published literature, workshops, patient-level practice data, patient self-completed questionnaire, patient-level provider data, ethnographic data and one-to-one interviews. Results Workstreams 1 and 2: intervention and trial successfully developed then refined. Workstream 3: glycated haemoglobin was not significantly different (p = 0.503) between intervention and control conditions (adjusted mean difference −0.10 mmol/mol, 95% confidence interval −0.38 to 0.18; −0.01%, 95% confidence interval −0.03% to 0.02%). Both patient-level referral to, and attendance at, structured self-management education programmes were lower or similar during the intervention than control conditions. There was no significant difference in most other secondary outcomes. Prespecified analyses indicated that glycated haemoglobin was statistically significantly lower (p = 0.004) among ethnic minority individuals during intervention than control conditions (−0.64 mmol/mol, 95% confidence interval −1.08 to −0.20; −0.06%, 95% confidence interval −0.10 to −0.02). This difference was not clinically significant and self-management programme attendance did not improve. Ethnography analyses found that the intervention’s attractiveness and usefulness were not self-evident to practices and providers, much of the activity was led by the embedders, and embedders covering multiple localities were not best placed to adapt the intervention to local contexts. Workstream 4: the intervention cost £0.52 per patient. There was no evidence of a difference in costs (−£33, 95% confidence interval −£2195 to +£2171) or quality-adjusted life-years (+0.002, 95% confidence interval −0.100 to +0.098) in the base-case analysis. The trial plus 12-month observational follow-up data showed that glycated haemoglobin was statistically significantly lower (−0.56 mmol/mol, 95% confidence interval −0.71 to −0.42; −0.05, 95% confidence interval −0.06% to −0.04%; p < 0.001) and self-management programme attendance higher (adjusted odds ratio 1.13, 95% confidence interval 1.02 to 1.25; p = 0.017) in intervention than control conditions, although it should be noted that the difference was not clinically significant. The qualitative substudy indicated that virtual programmes have a place in future self-management programme delivery, with highly positive feedback, particularly around financial and logistical benefits. Limitations The COVID-19 pandemic affected this research. A delayed start to the feasibility study prevented all learnings being taken into the wait-list trial, particularly around implementing the intervention at provider, not practice level. Practice engagement with the intervention was limited and variable. National Health Service commissioning restructures in England meant that, for many localities, changes to the provision of diabetes self-management programme commissioning included funding and capacity to co-ordinate and promote uptake in a similar way to the Embedding Package. With the wait-list design, a proxy primary outcome for self-management programme attendance was used, which may have affected the sensitivity of results. Finally, baseline structured self-management education programme attendance was higher than expected, and data sources were between 39% and 66% complete. Conclusions There were difficulties implementing the intervention, which probably contributed to the trial showing that, overall, the Embedding Package was unlikely to have affected glycated haemoglobin, self-management programme referrals and attendance or most other secondary outcomes. Future work Focus should be on which organisation(s)/role(s) can best drive change around embedding type 2 diabetes self-management programmes into routine care, and the role of blended face-to-face and virtual programmes. Trial registration This trial is registered as Current Controlled Trials ISRCTN23474120. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Programme Grants for Applied Research programme (NIHR award ref: RP-PG-1212-20004) and is published in full in Programme Grants for Applied Research; Vol. 13, No. 2. See the NIHR Funding and Awards website for further award information.
Plain language summary The problem: Self-management education and support programmes help people with type 2 diabetes to manage their diabetes. National Health Service guidelines recommend these programmes, but many people are not offered them and most do not attend. What we did: We tried to increase attendance at type 2 diabetes self-management programmes. We created an evidence-based ‘package’ of practical solutions, which included a website with useful resources and a person to encourage organisations to use these resources. We also created a patient digital self-management programme (MyDESMOND) as another way to receive support. We used the package in a small study and improved it based on what we learnt. We then carried out a large study to see whether the package worked and provided value-for-money. What we found: The package did not increase attendance at self-management programmes or improve patient glucose levels. There were many reasons for this; for example, many organisations did not use the package. However, there were small improvements in glucose levels among people from ethnic minority backgrounds, who are generally less likely to access self-management programmes. Also, glucose levels and programme attendance improved slightly when we looked at a longer period than the main study. The package was very low cost (52 pence per person). What does this mean? This package is a starting point for helping more people with type 2 diabetes access support to manage their diabetes, but more work is needed.
Scientific summary Background Type 2 diabetes mellitus (T2DM) remains a major health challenge in the United Kingdom. Structured self-management education (SSME) programmes help people with T2DM to manage their diabetes. These programmes are typically designed to engage people in developing and maintaining healthy habits (e.g. food, physical activity, medication taking, glucose monitoring) through peer and professional support and education. Evidence shows that SSME programmes are effective and cost-effective. T2DM SSME programmes are therefore recommended in national and international guidelines. Uptake of SSME programmes is poor, however, with the latest National Diabetes Audit figures showing that, within 12 months of diagnosis, 75% of people with T2DM were offered SSME but only 11% attended. Aim We aimed to develop and test an intervention (‘Embedding Package’) to increase T2DM SSME uptake by addressing barriers and supporting enablers to uptake at patient, healthcare professional and organisational levels. Objectives There were four workstreams. Public involvement underpinned the programme. Workstream 1: develop and tailor Embedding Package, including further development of a digital SSME programme (MyDESMOND); build capacity and develop resources; establish public involvement reference groups. Workstream 2: pilot Embedding Package to assess feasibility and suitability of components; Assess feasibility of collecting data with sufficient accuracy and completeness; refine Embedding Package. Workstream 3: conduct a wait-list cluster randomised controlled trial (RCT) with ethnographic study to compare and understand impact of Embedding Package with usual care on glycated haemoglobin (HbA1c) and SSME attendance [minimum clinically significant difference in HbA1c: 1.1 mmol/mol (0.1 %); target sample size: 66 practices]. Workstream 4: assess cost-effectiveness and sustainability of Embedding Package; conduct qualitative substudy to understand the impact of virtual SSME programmes. Methods Workstream 1: developing Embedding Package Mixed-methods development work for the Embedding Package, including: (1) secondary analysis of existing qualitative data, systematic literature review and the merging of the resulting coded data with further analysis; (2) stakeholder consultation through workshops and interviews; and (3) analysis of all collected data. This resulted in a list of evidence-based components needed by an intervention to increase SSME uptake. The multifaceted Embedding Package was then designed by a specialist education team based on this list. Workstream 2: feasibility study A single-arm, mixed-methods feasibility study in six practices across two Clinical Commissioning Groups (CCGs; East Midlands) piloted two patient recruitment approaches, data collection methods and intervention delivery in three participant groups: patients, practice staff, staff in SSME providers. Quantitative data were collected from primary care electronic medical records (extracted data, 2877 patients) and self-completed patient questionnaires (self-reported data, 423 participants). Health economics data were collected from SSME provider databases and questionnaires and interviews with practice managers and CCG staff. An integrated ethnographic study included observations, interviews and document analysis. Workstream 3: randomised controlled trial Using a wait-list cluster randomised design, with practice-level randomisation, we compared the Embedding Package with usual care (i.e. practices referred patients to SSME following their current standards). All practices provided usual care for 3 months (baseline, months −3–0). Practices were then randomised (1 : 1) to receive the Embedding Package for steps one and two of the RCT (months 1–18; immediate group; 33 practices; 16,340 patients) or for step two only (months 10–18; wait-list group; 31 practices; 18,393 patients). The primary outcome was patient-level HbA1c compared between intervention and control conditions as a proxy for SSME attendance, which was not suitable as a primary outcome due to the wait-list design. The main secondary outcomes were patient-level referral to and attendance at SSME (binary yes/no). Analysis used mixed models to account for patient- and practice-level clustering. Secondary analyses compared the completeness of SSME referral and attendance data from three sources (self-report, practice, SSME provider). The ethnography study used e-mail communications between the embedders (a key role introduced as part of the Embedding Package) and practices/providers, an intervention ‘tracker’ (database completed by the embedders to track embedding activities), embedder-generated documents, interviews with the embedders and one SSME provider, and observations of meetings between embedders and practices. Interpretive thematic analysis informed by normalisation process theory (NPT) was conducted. Workstream 4: assess cost-effectiveness and sustainability and enable implementation For the health economics, the primary (base-case) analysis estimated the costs of embedding activities as implemented across all practices within the RCT, and the discounted quality-adjusted life-years (QALYs) and costs in the intervention and control conditions. To assess sustainability, a 12-month observational follow-up took place immediately after the RCT during which the study team no longer actively reinforced the Embedding Package, but practices and providers could continue using it (minus the embedder). The RCT models for HbA1c, SSME referral and SSME attendance were repeated using the RCT data plus the observational data. The qualitative substudy (two primary care practices) comprised one-to-one telephone or video calls with practice staff (who were also virtual SSME educators) and people with T2DM who attended virtual SSME (‘attendees’). Topic schedules were mapped on the NPT and theoretical domains framework. Thematic analysis informed the data analysis. Intervention The Embedding Package comprised the key components of an intervention to embed SSME (see Results) delivered through a ‘toolkit’ and ‘embedder’. The toolkit was an online portal of resources, including tools, how-to guides and sample resources for anyone actively involved in implementing SSME (e.g. commissioners, providers, primary care staff). There were 88 live toolkit accounts, which could be accessed by multiple individuals. There were 19 page views on average per week for the toolkit (including by study staff as data could not be separated). The embedder also formed part of the intervention and was a new/amended role to work with practices and programme providers to embed support programmes into routine practice locally. The most and least common embedder activities were ‘promoting to patients’ (41% of activities) and ‘increasing referrals’ (3% of activities), respectively. Additionally, a patient digital support programme (MyDESMOND) was developed due to the absence of an online option at the time. The comparator was usual care. Results Workstream 1: developing the Embedding Package We developed the intervention as planned. Key components of an intervention to embed SSME into primary care were identified as: A clear marketing strategy for SSME. A user-friendly and effective referral pathway. New/amended professional roles. A toolkit of resources. Workstream 2: feasibility study The feasibility study showed that the RCT would be feasible, albeit with design improvements. Key findings included that the RCT primary outcome data (HbA1c) were over 90% complete in primary care, and 91% of patient-participants completing questionnaires consented to their responses being linked with their extracted primary care data. There was limited engagement from most participating practices and focusing the intervention implementation on practice staff was not feasible given capacity constraints in practices. It was therefore planned that, in the RCT, the focus of intervention implementation would be SSME providers. However, this proved difficult because RCT data were collected at the practice level, therefore embedders had to maintain a practice-level focus. Workstream 3: randomised controlled trial Baseline SSME attendance was higher than expected, at 64% and 38% in the wait-list and immediate groups, respectively. While providers were included in the RCT, much of the embedding activities remained focused at the practice level. The practices interacted minimally with both the embedder and the toolkit. Of the 66 RCT practices, 17 did not engage and 4 only used display boards in waiting areas. The remaining 45 practices had at least an initial meeting with an embedder. The primary outcome (HbA1c) was not significantly different (p = 0.503) between intervention and control conditions [adjusted mean difference, −0.10 mmol/mol, 95% confidence interval (CI) −0.38 to 0.18; −0.01%, 95% CI −0.03% to 0.02%]. There was also no significant difference in most of the secondary outcomes. Prespecified analyses indicated that HbA1c was statistically significantly lower (p = 0.004) among ethnic minority individuals during intervention than control conditions (adjusted mean difference −0.64 mmol/mol, 95% CI −1.08 to −0.20 or −0.06%, 95% CI −0.10% to −0.02%). This difference was not clinically significant and SSME attendance was not significantly different between intervention and control conditions. There was no difference in HbA1c among white individuals between the intervention and control periods. The three data sources for SSME referral and attendance data (practice, self-report and provider data) ranged from 39% to 66% complete. The ethnography component found that the attractiveness and usefulness of the Embedding Package were not self-evident to participating practices and providers, and most activity was led by the embedder, with much less active engagement from practice/provider staff. The analysis showed the importance of adapting the Embedding Package to local contexts, and that an embedder covering multiple localities was not necessarily best placed to undertake this. Workstream 4: assess cost-effectiveness and sustainability and enable implementation The pooled cost per patient of the Embedding Package was £0.52. There was no evidence of a difference in costs (−£33, 95% CI −£2195 to +£2171) or QALYs (+0.002, 95% CI −0.100 to +0.098) in the base-case analysis. The sustainability analyses (18-month trial data plus 12-month observational follow-up data) showed that HbA1c was statistically significantly lower (−0.56 mmol/mol, 95% CI −0.71 to −0.42, or −0.05%, 95% CI −0.06% to −0.04%; p < 0.001) and support programme attendance higher (adjusted odds ratio 1.13, 95% CI 1.02 to 1.25; p = 0.017) during intervention than control conditions. There was no statistically significant difference in SSME referrals. The qualitative substudy strongly indicated that virtual SSME programmes have a place in future education delivery. Feedback on virtual programmes was highly positive, particularly for the financial and logistical benefits. Negative perceptions included information technology challenges, a lack of interaction and informal conversations with fellow attendees, and difficulties in reading body language and visual cues during virtual delivery. Conclusions This programme was designed to address a gap in understanding how to improve uptake to T2DM SSME. We found that, although there was initial enthusiasm for the Embedding Package in some practices, there were difficulties in implementing it, and so many practices did not receive the intervention as it was intended. These were in part due to the embedder role sitting centrally with the study team, rather than in the locations where the package was being implemented or with the local provider organisation. Furthermore, since the proposal for this programme, the structure of NHS commissioning was reformed within England, which meant that, for many localities, changes to the provision of diabetes self-management programme in commissioning plans included funding and capacity to co-ordinate and promote uptake in a similar way to the Embedding Package. This limited uptake of the Embedding Package likely contributed to it having no overall effect on the primary or main secondary outcomes. It is also important to consider the impact of the COVID-19 pandemic, as this started during the latter step of the RCT and meant that all activities by the embedder were stopped prematurely. There were, nevertheless, some promising results. First, HbA1c (primary outcome) improved among ethnic minority groups during the periods when the Embedding Package was being implemented. However, recorded SSME attendance did not improve among this group, and so this may not have been due to the Embedding Package. Second, during the longer-term analyses, HbA1c was lower and SSME attendance higher during the intervention than control periods. This finding suggests that the positive impact of the Embedding Package may have taken longer than expected to realise highlighting that system-wide change such as this may require longer to be realised. However, other external factors could have also brought about this improvement. Implications for health care National roll-out of the Embedding Package in its current format is not suitable given the findings of this programme. However, given the low per-patient cost of the intervention and that the results were indicative of positive changes in the long term, we will instead create a PDF version of the toolkit so that the information contained within it can still be accessed by practices and SSME providers. The potential improvement in HbA1c among ethnic minority groups is of note. T2DM is known to disproportionately impact ethnic minority groups. It is therefore imperative that any initiatives to increase uptake to T2DM SSME do not further widen inequalities around access to support, which is the case with most initiatives in the UK and other industrialised countries. Conversely, implementation of the Embedding Package was associated with improvements in HbA1c in people in an ethnic minority population. This suggests that the Embedding Package, or parts of it, may have a role in future efforts to increase SSME uptake among ethnic minority groups. MyDESMOND became more important during the COVID-19 pandemic and is now among the top four diabetes applications (apps) on the ORCHA app library. Further evaluation is needed. The qualitative substudy findings suggest that patient choice could be important, particularly around having the option to attend SSME face-to-face and/or virtually. Key recommendations for research Consider how best to increase uptake to SSME within the new context of increased focus on digital/virtual modalities due in part to the COVID-19 pandemic, while maintaining the benefits to some underserved populations without alienating others and widening disparities (e.g. those without internet access). Consider how to tailor initiatives to embed SSME into routine care for underserved populations. Consider how best to embed SSME into routine care, particularly whether a local embedder working within a provider organisation is best placed to drive change. Evaluate blended approaches that combine face-to-face and virtual SSME delivery. Further evaluate MyDESMOND. Patient and public involvement The purpose of patient and public involvement was to involve people affected by the research and potential outcomes of the research at all levels of the project. The public involvement plan comprised five main areas of work: (1) public contributors involvement in study groups; (2) oversight and wider involvement through practice Patient Participant Groups (PPGs) and other local groups; (3) support via a local stakeholder group; (4) MyDESMOND development and consultation; and (5) dissemination of findings. Contributors provided governance and oversight of the project, offered advice on recruitment, publicity, local culture and dissemination, and acted as a critical friend giving feedback on project delivery in the local context. Public involvement was challenging because we needed to grow new networks in the localities involved in the RCT. The PPGs at participating primary care practices offered a potential solution to this that could be a useful approach in other research studies. Trial registration This trial is registered as Current Controlled Trials ISRCTN23474120. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Programme Grants for Applied Research programme (NIHR award ref: RP-PG-1212-20004) and is published in full in Programme Grants for Applied Research; Vol. 13, No. 2. See the NIHR Funding and Awards website for further award information. |
|---|---|
| ISSN: | 2050-4330 |