Algorithmic Approaches to Personalized Health Care

The U.S. health care system spends major resources on the treatment of acute conditions in a hospital setting rather than focusing on prevention and keeping patients out of the hospital. While there is no broad agreement on the potential solutions, structural reform is on the horizon. The meaningful use of Electronic Health Records (EHRs) is seen as a key to improving efficiency. Technology solutions including miniaturized implantable medical devices, networked home monitoring devices, and the ubiquitous use of smart phones are starting to enable continuous real-time monitoring of patients as they go about their daily lives. Rich data from these devices form an electronic Personal Health Record (PHR) that captures patient health in a much finer time-scale than the EHR. The health care system, however, is not well equipped to benefit from the impending deluge of personalized health-related data in order to improve health outcomes and reduce costs. This proposal puts forth a comprehensive and systematic approach to intelligently process such data aiming at preventing hospitalization, empowering patients to actively participate in managing their health, assessing quality of care, and facilitating cost-effective epidemiology in the emerging data-rich environment. In the proposed framework, early risk assessment starts with algorithms for mining EHR and PHR data to classify patients in terms of the risk they have for developing an acute condition that would require hospitalization and/or incur large costs. Risk stratification produced by this approach triggers a set of actions, including tests, additional and more intensive monitoring, and physician involvement. As part of this dynamic health management process, patients can have access to tools that enable their active participation in the daily management of chronic conditions, such as diabetes. Plans include the development of algorithms that leverage experts? opinions to assess quality of care and the development of a distributed epidemiology approach suitable for the emerging landscape where lots of data about each patient are distributed among many different locations.

The proposed work has the potential to achieve revolutionary improvements in the quality of health care. Risk assessment combined with intelligent management of chronic conditions can prevent acute health episodes and dramatically improve health outcomes. Having rigorous and scalable ways of assessing the quality of care has the potential to reduce medical errors and improve coordination among health care providers. On the educational front, plans include new courses, training a diverse set of graduate students, involving undergraduate students, actively collaborating with medical doctors, and reaching out to high school students through existing programs embraced by the PIs. Dissemination plans include capitalizing on the BU Sensor Network Consortium and organizing a major medical informatics workshop.