iSGTW Feature - Dan Reed at HealthGrid08

Feature - Biomedical informatics: Vision and challenges

Dan Reed
Image courtesy of Microsoft Corp.

Imagine your cell phone ringing at 6:00 a.m.: "Please telecommute today and keep children home from school. We are attempting to contain a sudden and dangerous outbreak of influenza that has already affected 30% of the population within ten miles of (your address)."

Dan Reed, a computational scientist known for his leadership in developing technology and science policy, currently director of scalable computing and multicore at Microsoft Research, envisions such a device in every pocket, hooked into a distributed data infrastructure. In his keynote address to the HealthGrid 2008 conference in Chicago on June 3, Reed invoked Vannevar Bush's post World War II vision of an "intelligence amplifying device." Admitting we're not there yet, he offered his vision of tools and practices to advance the field of public health informatics. He also laid out the challenges.

"The question is: How to get the right information to the right person at the right time?" Reed said. "Tools need to be context sensitive, dynamically adaptable. I see a future of embedded sensors and actuators picking up and transmitting biological and environmental data. With timely and personally relevant information at our fingertips, we can make better choices."

But he acknowledged there's a long way to go, and not just from a technical standpoint. Public health presents social challenges that may prove even more difficult to address than the technology. First, implementation of policies; for example, privacy: who can access a person's health record, and which parts? Second, investment and sustainability: once an infrastructure is in place, will funding continue year after year? Third, democratization: will researchers develop tools only the well-off can afford? The list goes on.

Reed argued that despite colliding interests, the health and information technology communities need to collaborate to develop solutions and to secure sustained support in the face of regulatory logjam. Tension will always exist between the research community and funding agencies; the challenge is to get researchers to approach policymakers with one voice.

"Technology doesn't convince policymakers," says Reed, "We need consensus. A confluence of disciplines has to agree on an approach and define standards."

Distributed data sources and their web services fusion are
changing our world. Click to see more.

Image courtesy of Microsoft Corp.

Given the mass of health data, he suggested changing from a hypothesis-driven to a data-driven approach. The challenge becomes one of data organization, to facilitate access across the domain, and formulating meaningful questions.

"Governments need to realize that they have to keep spending every year to keep infrastructure working and growing," Reed says. Referring to policymaker terms and election cycles, he adds, "The reward structure doesn't encourage this."

-Anne Heavey, iSGTW