23) There Is A 90% Probability That Your Son is Pregnant: Predicting the Future of Predictive Analytics in Healthcare

My Folder

Dale Sanders (Senior Vice President – Health Catalyst) Predictive: Relating to or having the effect of predicting an event or result. Analytics: The systematic computational analysis of data or statistics. Together they make up one of the most popular topics in healthcare today. But predictive analytics is a means to an ends, and there is little good in predicting an event or result without a strategy for acting upon that event, when it happens. If, as the Robert Wood Johnson Foundation recently published, 80% of healthcare determinants fall outside of the healthcare delivery system as we traditionally define it, should we focus our predictive analytics on the traditional 20% of traditional healthcare delivery, or broaden our focus to the 80% that includes social and economic factors, physical environment, and lifestyle behaviors? What if our predictive models reveal to us that the highest risk variable to a patient’s length of life and quality of life is their economic status? Can an accountable care organization and patient centered medical home realistically do anything to reduce that risk, in reaction to the predictive model? Given the current availability and type of data in the healthcare ecosystem, and our organizational ability or inability to realistically intervene, where should we focus our predictive and interventional risk management strategies? There is enormous potential value in the application of predictive analytics to healthcare, but, in contrast to predicting the weather, credit risk, consumer purchasing habits, or college dropout rates, the data collection, and social and ethical complexities of applying predictive analytics in healthcare are significantly higher. This session will explore some of the less technical, more human interest and philosophical issues, associated with predictive analytics in healthcare, including the speaker’s experience prior to healthcare, in the US Air Force, National Security Agency, and manufacturing.