Analysis of free text in electronic health records for identification of cancer patient trajectories

摘要: With an aging patient population and increasing complexity in disease trajectories, physicians are often met with complex histories from which clinical decisions must be made. Due to the rate of adverse events hospitals facing financial penalties for readmission, there has never been a greater need enforce evidence-led medical decision-making using available health care data. In present work, we studied cohort 7,741 patients, whom 4,080 were diagnosed cancer, surgically treated at University Hospital years 2004-2012. We have developed methodology that allows trajectories cancer patients estimated free text electronic records (EHRs). By these predict 80% ahead time. control confounders 8326 quantified events, identified 557 constitute high subsequent risks (risk > 20%), including six seven metastasis. believe presented findings could used improve decision support personalize thereby decreasing optimizing treatment.