Methodology
The overall concept of REDDIE is to identify from key stakeholders the methodological challenges to policymakers associated with new therapies (drug, devices, digital) for diabetes, and then to bring together four large, diverse real-world datasets to develop techniques to harmonise and develop synthetic data to create state-of-the-art models that policymakers can use to assess the impact of different variables on outcomes of an intervention. These can then be used to conduct “virtual trials” to supplement and support future RCTs or assess the benefits of novel devices or digital health interventions.
The REDDIE consortium has divided the project into a number of discrete work packages that address different aspects of the problem, but will work together in parallel towards the overall objectives of setting up standards for the use of RWD in assessing the efficacy, safety and value for money of interventions in diabetes. By engaging with a broad range of stakeholders, including members of regulatory authorities, payers, HTA boards as well as modellers, health economists, clinical trialists, industry and people with diabetes we will ensure that our deliverables and outcomes will be both relevant and impactful.
One of the main hurdles using RWD is to understand the difference in outcomes from interventions in RCTs and in RWD. Our approach will compare outcomes from RCTs with matched and unmatched populations from RWD databases to understand the efficacy to effectiveness gap and the factors that affect this.