Priority 1
The intervention had limited success and nationally SSME attendance remains fairly low at 11% within 12 months of T2DM diagnosis13 and 38–64% in the longer term (based on this study and Nicolucci et al.14). Research into embedding SSME therefore remains a priority and some of the original research questions remain, particularly:
- Which organisation(s) and role(s) can best drive change regarding SSME uptake?
- Would a local embedder and/or one within SSME provider(s) have more traction?
- Would integrating a behaviour change theory with an implementation theory/model into the intervention development and delivery be beneficial in addition to focusing on intervention content?
- How do direct-to-patient approaches (e.g. ability to self-refer) impact SSME attendance?
Future research should consider how initiatives around embedding SSME into routine care should be tailored for different populations, otherwise SSME referrals may inadvertently exclude some populations.
Priority 2
COVID-19 affected face-to-face, group-based SSME delivery, with greater emphasis on online modalities. This increased SSME access for some underserved populations (e.g. individuals who cannot take time off work for SSME). Future research should consider how to increase SSME uptake within this new context, maintaining the benefits to some underserved populations without alienating others (e.g. those without internet access).
Priority 3
Establishing reliable SSME referral and attendance data is key to obtaining reliable data for research to better evaluate these programmes and fully embed them within primary care.
Priority 4
Evaluation of virtual and blended (face-to-face and virtual) SSME delivery is needed.
Priority 5
Regarding public involvement, future research should continue to build methodology for engaging diverse communities in new locations close to where research is being conducted, rather than where investigators are based.
Priority 6
Further investigation would benefit wait-list design methodology challenges:
- Choice of primary outcome when the desired end point is terminal.
- How best to analyse terminal end points.
- Challenges around model fit when a high proportion of participants have only one observation across the steps.