Next Generation Informatics for Population Health Management and Measurement
We are experiencing a fundamentally new way human diseases can be studied, and health care can be delivered. This opportunity window is enabled by the vast amounts of patient data that are now collected across different digital modalities in and out of the clinic, coinciding with the maturation of computing machines and algorithms to efficiently analyze big data. Although tremendously exciting, leveraging new types of health and medical datasets in clinical decision support and for new knowledge discovery is not without substantial challenges. In this talk, I will describe ongoing efforts to improve outcomes for patients at risk of suicide and overdose events at the Veterans Health Administration using next generation informatics tools for predictive analytics and natural language processing of patient records. I will also highlight real-world challenges I have experienced designing, deploying and maintaining such tools within the largest integrated delivery system in the US.