Table 8 Summary of findings for the Decision-making after Pilot and feasibility Trials (ADePT) process [28]

1. Did the study allow a sample size calculation for the definitive trial?

4

Using HbA1c as a primary outcome, the sample size for future randomised controlled trial was estimated to be 492 participants (41 participants from 12 clusters).

2. What factors influenced eligibility and what proportion of those approached were eligible?

1

All diabetes centres approached were eligible. In order for young adult participants to be eligible, they had to meet the inclusion criteria and also have attended a clinic appointment between Oct 2019 and Jan 2020.

3. Was recruitment successful?

1

Recruitment proceeded smoothly in 2 of the 4 centres. Due to differing ethics requirements, there were more barriers to recruitment which led to lower numbers of participants being recruited in the remaining 2 centres.

4. Did eligible participants consent?

1

Consent was obtained successfully in all centres

5. Were participants successfully randomised and did randomisation yield equality in groups?

1

This was a cluster randomised RCT. Randomisation was successful. However, one centre from the control arm left the study due to delays with ethics and the onset of COVID-19. This, along with additional ethics committee requirements around recruitment, meant the numbers of participants in the control arm was small (see CONSORT flow diagram for detail).

6. Were blinding procedures adequate?

1

Not applicable to current study

7. Did participants adhere to the intervention?

1

Young adult participants and support workers filled in the “study delivery checklist” after each appointment. Adherence was generally good. Three consistent deviations were identified: see text for details

8. Was the intervention acceptable to participants?

1

Young adult participants’ perceptions of the agenda-setting tool and support worker were very positive. Their views on Florence were mixed. Staff held similar views on the intervention components but had an additional concern around resourcing of the Support Worker.

9. Was it possible to calculate intervention costs and duration?

3

The implementation cost of the D1 Now intervention over 12 months was estimated at €1281 per young adult for the internal support worker arm and €1804 for the external support worker arm.

10. Were outcome assessments completed?

2

Young adult self-reported outcomes were 100% completed at baseline and 71% at time 2. This second lower rate of completion was likely caused by time 2 data collection happening through the post, rather than in the clinic, due to COVID-19 related restrictions.

Young adult clinical outcomes were 100% completed at baseline and time 2, with the exception of HbA1c, which again was impacted by COVID-19 related restrictions.

11. Were outcomes measured the most appropriate outcomes?

2

Yes, outcomes were based on a core outcome set developed for this population [22]. Young adult participants were extremely positive about being asked to complete questionnaires dealing with the psychosocial aspects of living with diabetes.

12. Was retention to the study good?

1

Yes, there was just 12% attrition from baseline to follow-up. Details are in the CONSORT flow diagram.

13. Were the logistics of running a multicentre trial assessed?

1

Yes. Patient recruitment and intervention delivery in a future definitive trial was identified as being resource intensive. The varying requirements of multiple different Research Ethics Committees were also identified as a challenge.

14. Did all components of the protocol work together?

1

Yes, but some modifications are required to move forward to a definitive RCT.