We aimed to enrich eHOP, the big data platform of the French CDC Network (Ri-CDC) with cancer specific data sources (including omics and tumor board reports) to conduct automatic patient pre-screening. We present here the main difficulties to address and our preliminary results: extraction and representation of eligibility criteria in clinical trials, entity extraction from electronic health records (EHR) and export to standard data model for clinical data reuse.
Learning Objective 1: Learn challenges in automation of pre-screening studies and possible solutions, including variables involved in eligibility criteria, how to extract them from free-text and how to export them in a standard data model suitable to run eligibility queries.
Jeremy Pasco (Presenter)
CHRU of Tours
Boris Campillo-Gimenez, INSERM
Leslie Grammatico-Guillon, CHRU of Tours
Marc Cuggia, INSERM