GENz-trials

Clinical trials are seen as the gold standard for evaluating the efficacy, safety, and cost-effectiveness of many medical products including medicines, medical devices and diagnostic tests. However, setting up and running a clinical trial is far from easy. Identifying and recruiting patients represents a major barrier, and almost a third of trials are terminated early because of recruitment problems. Furthermore, around a third of patients who sign up for trials drop out before trial completion. There are also ethical questions surrounding the use of placebos; while essential for scientific rigour, they are hard to justify in trials involving chronic or life-threatening conditions.

The aim of GENz-trials is to harness the power of data science and artificial intelligence (AI) to address these challenges and transform the way clinical trials are designed, carried out, and evaluated.

Specifically,  the project plans to improve patient recruitment by using AI to analyse electronic health records (EHRs) and real-world data (RWD)  helping to better  match patients with suitable clinical trials. Another project goal is to use generative AI to automate key workflows, including drafting protocols, personalising informed consent discussions, monitoring results, and preparing for regulatory submission.

Finally, to reduce the numbers of patients receiving a placebo during clinical trials, GENz-trials will explore the use of RWD to construct digital twins and synthetic control arms (SCAs) which could at least partially replace the control arms in trials.

The project will start by drawing on real-world and clinical trial data to create AI tools for protocol generation, literature synthesis, consent, regulatory compliance, patient recruitment, and synthetic control arms based on actual cohorts. Next, they will test the tools by emulating real clinical trials, comparing the real experimental arm with both the real, cohort-based control arm and a control arm based on digital twins.

Finally, the full system will be validated in a real-world study comparing a generic anticoagulant with its branded counterpart. Here, the project will use SCAs based on digital twins to generate high quality evidence on effectiveness and safety.

The GENz-trials tools should improve clinical trials in multiple ways: by facilitating recruitment, cutting the administrative workload, and reducing the need for large control groups, in order to speed up trials and cut costs. Furthermore, the synthetic control arms will make trials more ethical.

Ultimately, the GENz-trials team hopes their outputs will boost innovation in the assessment of healthcare solutions, advancing the evaluation of not only pharmaceuticals, but diagnostics, digital health tools, and therapeutic combinations by regulators, health technology assessment bodies, and payers. In the long term, this will speed up patient access to much-needed healthcare innovations.