Table 3.

Key points from decide trial

Trial name (registration)	DECIDE: Pragmatic Randomised 104 Week Multicentre Trial to Evaluate the Comparative Effectiveness of dapagliflozin and Standard of Care in Type-2 Diabetes17 (NCT02616666)
Sponsorship	Industry then academic – AstraZeneca then Liverpool.
Clinical setting	Type 2 diabetes mellitus on metformin and needing better glucose control in primary care setting, UK only: national.
Design	Phase 4, randomised, open label, comparator trial.
Size	N>800 (ongoing).
Key outcome measure	Clinical success (glucose control) at 12 months.
Rationale for presentation	Recruitment.
Key discussion points	DECIDE is a clinical effectiveness trial comparing dapagliflozin, a sodium-glucose cotransporter-2 (SGLT2) inhibitor, to standard care (alternative medications such as sulfonylureas and DPP4 (Dipeptidyl-peptidase 4) inhibitors) in people with type 2 diabetes requiring second-line pharmacological treatment (after metformin). There was a limited window in which patients need to be approached before treatment is changed. CPRD managed the trial using their interventional research services platform that supported general practitioners (GPs) and patients logging into the system. Algorithms were developed to carry out daily searches of primary care EHR for potentially eligible patients meeting prespecified criteria. The primary care team then reviewed the search list and enrol eligible patients into the study at their next routine scheduled visit. Most of the data required for the trial was available in CPRD’s routinely collected dataset, thereby facilitating collection of primary and secondary outcome measures as well as identification and recruitment. Patient-reported outcome measures were collected electronically. Trial monitoring could therefore be carried out centrally. Lessons from the trial are that EHR search algorithms are easy to set up, but feasibility assessments should take account of variability of coding and real-world clinical practice. Recruiting patients via EHR is achievable and allows easy collection of long-term data for follow-up, but as with all trials, it depends on GP and patient participation and engagement. Data enablement and platform technology allows real-time adaptation to real-time conditions.

CPRD, Clinical Practice Research Datalink; EHRs, electronic healthcare records.