A Computing Research Perspective on a Learning Healthcare System. Kevin Sullivan Computer Science University of Virginia 4/11/2013

Transcription

1 A Computing Research Perspective on a Learning Healthcare System Kevin Sullivan Computer Science University of Virginia 4/11/2013

2 Outline Motivation unmet potential, pressing need Goal use driven, fundamental research agenda Challenge it s a ULS systems problem Vision healthcare as a smart learning city Evidence we ve seen such emergence before Brainstorming research challenges in CISE Opportunity new community, deep impact

3 Motivation Revolution in information technology promises transformational improvements in healthcare E.g., Data to decisions Cyber physical systems Platform architectures Assurance technology Potential still untapped; progress far too slow Traditional approaches not working very well Why? How can we make the next LHS happen?

4 Goal Question here isn t what IT to build But to define research to greatly improve how our health ecosystem computes answers to key questions in health science situation awareness at all scales best courses of action for individuals, institutions whether to intervene in ongoing care processes how to improve ecosystem performance itself

5 Challenge National scale health system is an ultra largescale system, in sense of SEI report by this title Billions of lines of evolving code across ecosystem Billions of ever evolving computing devices Complex and conflicting goals and requirements People/institutions integral components of system 10^8 people; many thousands of organizations Legal /regulatory frameworks are architectural Domain deep in professional knowledge, practice Traditional top down engineering won t work State of art inadequate to needs of such a system

6 Vision Buildings/engineering versus cities/planning Buildings based on planning and top down control Cities too complex to be not built this way Yet good ones are far from being free for alls Planning, infrastructure, governance => emergence National LHS isn t a building, but a city a ULS system Top down engineering necessary but insufficient Develop science and technology to use Infrastructure, governance, value to drive planned emergence in ultra large scale systems, for healthcare, in particular

7 Evidence Internet TCP/IP, routers, IETF, value WWW, Amazon, Google, etc Apple Architecture, infrastructure, governance, value App Store phenomenon Question What advances needed to do this for healthcare? Catalytic combination of design, technology, governance, value How to make healthcare ecosystem a smart, learning city?

8 Brainstorming Efficient and effective data quality at U.L. scale Hybrid digital physical human institutional computing Requirements for ULS health ecosystem infrastructure Reliability, availability, and data consistency at scale Evidence driven planning, engineering, and assurance Hypothesis driven automated information pull Cyber social computing systems architecture How can IT change economic games decisively Multiparty computing without trusted 3 rd parties Big data, cyber physical systems, software eng., etc.

9 Opportunity CISE community can become engaged with the hard and important problems in national healthcare ecosystem Health community can become engaged at cutting edge of computer & information science & engineering Extraordinary opportunities for use driven, fundamental research in CISE, with huge potential for national impact Great potential to vastly accelerate IT catalyzed advances in health, health knowledge and practice, and economy

10 Conclusion No bigger problem than health and healthcare A community and agenda to break the logjam Far reaching vision and catalytic path forward Important, use inspired, fundamental research Leading to real infrastructure, governance, etc Ultimately transforming and vastly improving how the US healthcare ecosystem computes in the interest of its many and varied stakeholders