Oracle and AstraZeneca agree non-binding AI pact to find patients for clinical trials

Oracle and AstraZeneca on 20 October announced a non-binding agreement to test how Oracle's real-world data, cloud and AI tools can help clinicians find patients who qualify for clinical trials and feed research findings back into routine care¹. No financial terms were disclosed¹.

What the companies agreed

The agreement covers exploration of selected use cases rather than a committed programme¹. Oracle brings its health data, cloud, AI, analytics and clinical workflow software, while AstraZeneca contributes scientific and therapeutic expertise¹. The companies name patient recruitment as the problem they are targeting first, describing it as one of the largest obstacles in clinical trials¹.

Disease areas that could be explored at the start are listed as:

• oncology¹
• cardiovascular, renal and metabolic diseases¹
• respiratory and immunology¹
• rare diseases¹

Beyond recruitment, the two plan to try a continuous learning model in which results from trials and routine care inform later study design, safety monitoring and treatment¹. They intend to set up shared governance, reusable technical components and agreed outcome measures, and to use those measures to decide whether approaches that work should be extended to other disease areas and geographies¹.

AstraZeneca chief executive Pascal Soriot said the aim was to help clinicians "identify patients sooner, optimize clinical trials, and generate evidence continuously across the lifecycle of a medicine"¹. Seema Verma, who runs Oracle Health and Life Sciences, said the work would sit "under the governance of the health systems delivering care"¹.

What it means for the market

The deal pairs one of the largest drugmakers with a technology group whose health business sits inside clinical workflows¹. For pharma sponsors, the pitch is earlier and easier identification of eligible patients through the systems care teams already use, which goes to the cost and time of recruitment¹. For health systems, the release presents them as the gatekeepers of the data and the governance, while also offering their patients wider access to trials¹.

For vendors selling trial-matching and real-world evidence tools to pharma, the announcement signals that large sponsors are looking to work directly with the companies that own clinical workflow software. How far that changes buying is not yet clear. The release gives no budget, no named health systems, no patient numbers and no timeline¹. Whether any use case becomes a contracted service depends on the outcome measures the companies have yet to agree¹.

Caveats

The agreement is non-binding and its use cases are described as exploratory¹. The disease areas are ones that "could" be explored, not confirmed programmes¹. The only source is the companies' own announcement, issued from Austin, Texas and Abu Dhabi¹. Any extension across disease areas and geographies will turn on the shared outcome measures the two companies plan to set¹.