Biomedical Informatics Calendar

Over the last few decades, growing use of Electronic Medical Records (EMRs) systems have established large practice-based clinical datasets, which are emerging as valuable resources for clinical and translational research. The longitudinal nature of the data contained within EMRs makes them ideal for quantifying drug outcomes (both efficacy and toxicity). The major challenge of using EMR for studying drug outcomes has clearly been the identification of patients’ drug exposure and outcome information from heterogeneous clinical data, including narrative reports. This presentation will describe our recent development of natural language processing (NLP) and data mining methods for extracting and modeling medication data in EMR, as well as how such informatics methods can be used to supportongoing drug studies at Vanderbilt. Two EMR-based drug studies will be discussed: 1) a drug-repositioning study – to validate that metformin, a drug for diabetes, can increase cancer survival rate; and 2) a pharmacogenetic study – to replicate known genetic associations of warfarin stable dose.