Oracle Provides Buck Institute, Which Studies Aging, with Access to 122+ Million Medical Records for THRIVE, a Federal Aging Measurement Program

Oracle Provides Buck Institute, Which Studies Aging, with Access to 122+ Million Medical Records for THRIVE, a Federal Aging Measurement Program

On September 23, Oracle and Buck Institute for Research on Aging announced a partnership. The institute will receive access to deidentified longitudinal patient records and to the Oracle Life Sciences Data Intelligence analytics environment. These data will form the foundation of the THRIVE consortium's work on the Intrinsic Capacity aging measure.

Trials of interventions against aging typically stretch across decades: researchers must wait to see which participants go on to develop disability, serious chronic disease, or death, and which do not. ARPA-H, the agency that funds high-risk medical research, launched a program with a different aim, to learn how to predict such outcomes in roughly three years. THRIVE, a grant under that program awarded to a Stanford University team, is building a single Intrinsic Capacity measure for this purpose. It combines five dimensions of a person's ability to live independently (locomotion, cognition, psychological state, sensory function, and vitality) into one score.

The measure is useful for trials only if its early changes predict what will happen to a person years later. Proving this requires the longitudinal histories of real people over many years, and that is precisely what the Oracle partnership provides.

Buck will gain access to 122+ million deidentified longitudinal medical records from Oracle Health Real-World Data. The figure is as of May 2026 and counts unique patients within each health system separately; the same person, if treated in different systems, may be counted twice. With this dataset, Buck can assemble cohorts of patients with similar histories, trace their records over years, and look for early changes that are more often followed by frailty (rapid loss of physical capacity and independence) or chronic disease.

The ARPA-H consensus document on Intrinsic Capacity lays out this validation step by step: first, confirm the domain scores on large longitudinal datasets; then, in clinical trials, show that interventions change those scores and that the changes are linked to meaningful outcomes such as hospitalizations, loss of independence, and mortality. The 122+ million Oracle records address exactly the first step: they make it possible to test which combinations of features in a medical history actually precede its deterioration.

David Furman, a researcher at Buck, explained that access to long patient histories will let the team "begin linking early clinical markers to long-term outcomes." Oracle is providing Buck with hosted access to its Life Sciences Data Intelligence environment, where researchers build cohorts, run queries, and conduct bioinformatic analyses without having to prepare each dataset separately.

In February, THRIVE received up to $34.5 million from ARPA-H to develop the measure itself. The Oracle partnership supplies the material for its validation: years of already accumulated medical histories against which an early signal can be compared with what happened to a person afterward.