Embedding Package
Our findings highlight how knowledge gaps can arise when logistical barriers prevent the study design from reflecting intended real-world delivery.
While our findings do not support national roll-out of the intervention, providers and practices engaged well with parts of it (e.g. ‘how-to’ guides and ‘promoting to patients’) and the intervention was very low cost per patient. Furthermore, the national DESMOND Programme Office found that an adapted version of the toolkit was well accessed by their providers, signalling that a SSME programme-specific toolkit may have more real-world utility. Finally, similar toolkits are available in PDF format.40,41 We therefore developed a PDF format for initial intervention roll-out (see Report Supplementary Material 1), with the same content as the online toolkit and an added focus on online SSME delivery. The PDF toolkit was made available free-of-charge to participating practices and via the DESMOND website.
Online self-management programmes
A core part of the original proposal was developing a digital SSME programme because, at the time, no such programmes were available. MyDESMOND became more important during the COVID-19 pandemic and is now among the top four diabetes applications on the ORCHA application library. Further evaluation is needed.
Real-time, group-based, virtual SSME delivery by an educator was also shown as acceptable to people with T2DM and delivery staff. Virtual SSME delivery was viewed as key going forward to ensure that people with T2DM have choices around accessing SSME, and to increase availability.
Ethnic minority groups
The Embedding Package was associated with a small HbA1c reduction among ethnic minority populations, which cannot necessarily be attributed to the intervention. Nevertheless, this potential improvement is notable and may relate to activities focused on ethnic minority populations. For example, the embedder attended a gathering at a Hindu temple and delivered a presentation on diabetes and SSME, with attendees encouraged to self-refer on the day. A local dietitian answered questions, and the embedder sought volunteers to work with the local provider to adapt their SSME programme to ensure cultural appropriateness. Unfortunately, the pandemic stopped this work.
Type 2 diabetes mellitus disproportionately impacts ethnic minority groups,42 so it is imperative that strategies to increase SSME uptake do not further widen inequalities around accessing support. This is the case with most UK-based initiatives because, for example, majority ethnic populations may be more able to navigate systems or have English as their first language. Conversely, the intervention seemed to either benefit people in ethnic minority populations or at least did not increase existing inequalities.
The findings suggest that the Embedding Package, or parts of it, could be usefully applied to explore how to increase SSME uptake among underserved populations, particularly ethnic minority groups.
Other lessons learnt
COVID-19 affected the programme’s practical delivery and research focus and brought into stark focus the ability to adapt in a long programme of work.
A strong evidence base for an intervention does not always translate into effectiveness. Therefore, while interventions should be based on best possible evidence, robust testing is necessary.
To really understand the effectiveness of initiatives to increase SSME uptake, work is needed to improve SSME data quality, availability and linkage, which is currently of moderate standard.
Finally, implementing change around SSME uptake takes time to change stakeholder mindsets and embed SSME as part of routine care. While ambitious, with content grounded in comprehensive research, the Embedding Package did not draw upon theories of change management for systems or of behaviour change, which could be the next step towards creating real change.