Grant: Foundation Grant

To monitor how our health care system functions, researchers use data routinely collected through public services, such as bills from physician visits, hospital discharge information, and social data from non-health services. Analyzing these records allows us to examine factors that influence the health of Canadians and how we use public services. Until now, researchers have performed these analyses using structured data (such as spreadsheets) that can be easily plugged into statistical programs. However, with the development of new analysis techniques, there is great potential to conduct more in-depth research using unstructured data, such as clinical notes from electronic medical records (EMR). Big data analytics can be used to analyze free text (i.e. natural language). This proposal will employ these new techniques to analyze free text in the health and social data at the Manitoba Centre for Health Policy (MCHP). Big data analytics will allow us to answer broader research questions that will help to improve clinical practice. For example, Canadians sometimes receive unnecessary medications (antibiotics), and undergo unnecessary tests like back x-rays. By developing the tools to analyze clinician's notes (free text) in EMRs, we can examine why these events occur and better understand the role of patients in decision making. Using the extensive person-level data available at MCHP, we will follow patient exposure to antibiotics over time while considering travel history and other contributing factors. Thus, the inclusion of free text analysis adds an additional dimension to our understanding of antibiotic resistance and will help to curb this growing problem. This research will advance the science of population level data analysis to a new level, adding depth to the research questions we can answer, and supporting the development of evidence-based policy.