Oracle and AstraZeneca set out AI clinical research partnership

Oracle and AstraZeneca announced a collaboration on October 20 to explore how artificial intelligence and information generated in routine healthcare could connect clinical research with patient care. The proposed work combines the drugmaker’s scientific expertise with Oracle’s health data, cloud infrastructure and clinical technology. Its objectives include finding people who may qualify for studies and bringing research findings into the workflows clinicians use in everyday practice.

The companies have signed a non-binding term sheet, and AstraZeneca said financial details are not being disclosed. They envisage projects in the United States, the Middle East and Europe, working with local governments and following applicable regulations. The disease areas under consideration span cancer, cardiovascular, kidney and metabolic conditions, respiratory and immunological diseases, and rare diseases. The announcements describe an exploratory collaboration, with individual projects still subject to their own review.

Patient recruitment is one of the problems the partners want to address. They plan to examine whether AI-supported applications can help care teams recognise potentially eligible people earlier, making research opportunities easier to identify within clinical work. AstraZeneca also identified earlier diagnosis, monitoring the safety of medicines and using real-world evidence throughout a medicine’s lifecycle as areas for investigation. These are proposed applications; the announcements do not present measured improvements in diagnosis, recruitment or treatment outcomes.

AstraZeneca said the collaboration would use deidentified real-world data under privacy, security and governance arrangements, and that it would not receive information that identifies individual patients. Each proposed project will be assessed separately and proceed only when the relevant legal, privacy, regulatory, ethical and institutional safeguards are in place. Oracle also placed the planned work under the governance of the health systems providing care.

The broader research model would allow findings from studies and routine care to inform later study design, safety monitoring and clinical practice. Oracle said the companies plan to agree on shared governance, reusable technical building blocks and measures for assessing outcomes. Those measures would help determine whether an approach that works in one setting could be extended to other conditions or regions. The arrangement therefore includes evaluating results before deciding on wider use.

The first programme is planned for Abu Dhabi with healthcare company M42. Announced at the Abu Dhabi Future Health Summit, that initiative will explore support for people whose hypertension remains uncontrolled. M42 brings local care delivery and research experience, AstraZeneca contributes cardiovascular and clinical development expertise, and Oracle supplies electronic health record, cloud, AI and workflow capabilities. The proposed tools would use approved clinical information already available to care teams to flag patients for further assessment or treatment review.

The Abu Dhabi work also covers identifying people who might meet initial eligibility criteria for selected clinical trials. M42 said authorised research teams would use the proposed approach, with qualified professionals reviewing potential matches. Contacting patients or enrolling them would follow approved study procedures and documented informed consent. Healthcare professionals would retain responsibility for decisions affecting patients, while the care element would seek to support treatment aligned with clinical guidelines.

The programme follows an earlier expansion of M42’s relationship with Oracle Health. In September, M42 announced plans to adopt Oracle Health Data Intelligence and integrate it with Oracle Health Foundation EHR, connecting clinical, genomic and real-world information across its network. That separate announcement named diabetes programmes as the intended starting point for Data Intelligence.