Cytel applied machine learning to find patient subgroups that benefit from treatment after a severe infection trial missed its primary endpoint.

In a double-blind study in severe infectious disease the overall population failed to meet the primary endpoint, and simple sub-group analyses could not show which patients benefited versus standard of care. Cytel applied machine learning methods to identify patient subsets that could benefit from the active treatment. The page gives no quantified results.

Cytel is a biostatistics and clinical trial design firm whose East Horizon software is used by pharma sponsors for adaptive design, simulation and regulatory-ready statistical analysis

This sits in the Build stage of the AI use-case lifecycle: Named examples from the foundation-model labs and platforms pharma teams use to actually construct, train and generate the model or application.