Increasing the Patient-Centeredness of Predictive Analytics Tools

Excerpt

Predictive analytics tools are fundamentally transforming the delivery of healthcare and the practice of medicine, but not necessarily in ways that matter to patients. Predictive analytics tools use artificial intelligence/machine learning (AI/ML) and other advanced statistical methods to identify patterns and predictions about a multitude of health outcomes. Perhaps the best-known application of predictive analytics in medicine is the use of automated risk factors in the electronic health record, which alerts physicians to screening and treatment priorities [1]. An emerging class of predictive analytics tools is intended to predict patient preferences for complex and challenging treatment decisions, such as predicting the end-of-life care wishes of incapacitated patients [2]. Predictive analytics tools aspire not only to aid clinical decision making but also to surpass the accuracy of human decision making, reduce errors, and lower costs. …