The 23 March webinar featured input from GetReal’s Work Package 1 (WP1): Sarah Garner, Associate Director Science Policy and Research, National Institute for Health and Care Excellence (NICE); Pall Jonsson, Senior Scientific Adviser, NICE; and Rob Thwaites, Senior Director, Takeda. In addition to providing an update on progress with the GetReal project, the webinar introduced the framework that is being developed for use by all parties to help plan for assessing relative effectiveness of new therapies.
Sarah Garner began by setting the scene with an explanation of the effectiveness issue. Garner first pointed to the “traditional” paradigm of drug development, based on a “hierarchy of evidence” rooted in the traditional Randomised Control Trial (RCT). When HTA bodies entered the picture, people stated asking more questions about how new medicines were stacking up in the real world. Thing started to change, with people questioning if RCTs are indeed the best model for new drug development. Increasingly important became the topic of “effectiveness” –“evidence used for decision-making that is not collected in conventional RCTs”.
While studies rooted in effectiveness allow for examination of patient benefit and harm in a context in which the technology is applied in everyday practice, these studies also tend to result in more “dirty” data, noted Garner, largely due to variability and biases. Challenges described by Garner include:
- Phase III trials are too short to capture relevant effects and need to use models, leading to considerable uncertainty in RWE predictions;
- RWE is likely to be influenced by factors (i.e. adherence) that are not captured in Phase III, while model-based estimates unreliable;
- Phase III trial event rates for comparator are unlikely to be in line with available RW evidence for the comparator, suggest a RWE bias.
A full list of “effectiveness challenges” described by Garner can be found in slide 9 the webinar slide pack.
Speaking after Garner was NICE’s Pall Jonsson, who elaborated on just how IMI GetReal has been working to address the efficacy-effectiveness gap. Improved understanding of the efficacy-effectiveness gap can better inform just when and how real world evidence should be used, Jonsson explained—information that can in turn help improve drug development strategies.
Towards this end, a key initiative of GetReal’s WP 1—which focuses on framework, processes and policies—has been stakeholder engagement. The project has included a series of workshops pulling together stakeholders to discuss issues and possible options for redesigning the development strategy.
One part of these discussions centred on case studies in specific disease areas, including non-small cell lung cancer, rheumatoid arthritis, and COPD, among others. In the COPD case study, for instance, the group looked at early pragmatic clinical trials for measuring effectiveness. Questions considered by the stakeholders included:
- When and why are these types of studies useful?
- What are the barriers to stakeholder acceptance of early pragmatic trial data?
- What are the barriers to acceptability for decision-makers?
The key point in the examination, noted Jonsson, was not the different disease areas themselves, but rather the different analytical techniques and study designs within these disease areas.
Providing the industry perspective, Rob Thwaites of Takeda introduced the real-world evidence framework that GetReal is working towards. The framework outlines different means of addressing various effectiveness challenges, compiling this information in an online platform that will be available to everybody, both those who are generating evidence and those who are assessing evidence.
One example cited by Thwaites was the issue that RWE is likely to be influenced by factors such as adherence, which are not captured in Phase III model-based estimates. Towards this end, GetReal is focusing on options like pragmatic clinical trials, which could be initiated prior to approval. A revised pathway could see the introduction of a pragmatic P3b trial: