Researchers said they were able to more than double, on average, the pool of eligible patients in oncology clinical trials by using artificial intelligence (AI) to broaden restrictive eligibility criteria.

In a retrospective study, researchers from Stanford University, in partnership with Genentech, used electronic health records (EHR) of 61,094 patients with advanced nonsmall-cell lung cancer. The data, which came from a nationwide database of highly structured EHRs from Flatiron Health, was then analyzed by Trial Pathfinder, an AI framework that uses data-driven algorithms and real-world data. The results were that many common criteria, including exclusions based on several laboratory values, had a minimal effect on trial outcomes.

The results were published online by the peer-reviewed journal Nature on April 7.

Read the report here: https://go.nature.com/3g0rCDM.