Jennifer Zaino
SemanticWeb.com Contributor

Partners Health Care knows that its future competitiveness and ability to manage care in a new era of genetics and genomics depends on having a solid, clinical knowledge-management strategy.

Toward that end, the integrated non-profit health care system — whose members include the Brigham and Women’s Hospital and Massachusetts General Hospital academic medical center, as well as a host of specialty facilities, community health centers, and other health-related entities — is embracing semantic web technology. It is putting in the tools and the infrastructure that will let it keep up with the speed of information change in the medical arena with consistency across its entities, while using the minimum amount of labor.

That’s a job for the semantic web.

“In everything related to knowledge representation, the maintenance of content, anything related to tools, editors, and content management lifecycle — that is the role of semantic web technology,” says Dr. Tonya Hongsermeier, corporate manager for clinical knowledge management.

Hongsermeier’s team is responsible for the whole knowledge-management enchilada, establishing the tools and infrastructure for everything ranging from clinical decision support to decision capture to content lifecycle management to editing and content delivery systems. Every day she lives the challenge of restructuring knowledge so that it is maximally shareable and reusable from a services perspective, where data can be created in one context and reused in another through metadata extensions. And if you think it’s tough to do that today given the ever-expanding and ever-changing base of medical knowledge, the future is going to present even more challenges.

“Genetics and genomics will cause another level of knowledge explosion, in the volume and velocity of knowledge processing,” says Hongsermeier. “The only hope to meaningfully provide good guidance on how to use genetic information is through federation activities, whereby knowledge is created in a variety of places and context, but semantic web technologies let us federate those knowledge aspects and expose them, using SOA models, to decision support stakeholders. Knowledge reuse, maintenance and so forth are dramatically impacted by semantic web technologies.”

In health care decision support environments, the ability of a doctor to get an asset update as a byproduct of being a customer of that asset, enabled by the notion of description logic and OWL (Web Ontology Language), is pretty powerful. That’s a radical change from how things are typically done today, when knowledge is cloned and reproduced in 5,000 places in a database-but the problem goes beyond that.

“We could have 5,000 rules with the same knowledge components in 5,000 places. We also could have the same knowledge in the rules as an artifact in the clinical documentation system, as a report in a SQL data warehouse, as an order set,” says Hongsermeier. “So the problem is that it’s not just horizontal across 5,000 points of light in the rules, but there are also different types of knowledge artifacts that may need to consume that same piece of knowledge.”

For example, today the dataset about why a patient can’t take the drug Warfarin is reproduced in rules; in documentation forms that surface to ask if there is a reason a patient doesn’t take Warfarin; and in reports about whether anyone in a health care system received Warfarin that shouldn’t.

“Semantic web technologies let you externalize that definition-contraindications to Warfarin. Rather than instantiating that in the rule, the rule points to it, or the form points to it as an ontological object, rendered as an RDF glob of data that everyone can point to,” she says. “So, whenever we ask the physician about this issue, we pull that knowledge object and surface it in this rule and surface it in this report, and so on.”

But there remain obstacles to Hongsermeier’s mission to fully leverage the semantic web to help Partners realize its knowledge management strategy.

“I think that we need the commercial grade tools to make a huge amount of progress,” she says. “I cannot buy a commercial grade web ontology language editing environment or a run time environment. Basically, the community is still a lot of very technical folks tinkering with tools that only people with PhDs can use. We need to move towards tools, infrastructure and products that business level people can use — that are scalable and implementable in large organizations. That’s the tipping point we are looking for.”

While there is a lot of open source technology available, Hongsermeier doesn’t feel confident putting in a commercial grade environment. Especially not a health-care setting, where people’s well-being is at stake. “We have to have trust that when we edit knowledge and deploy it, that it’s right, consistent with specs, and that it won’t have something in it that would hurt a patient because something was lost in translation.”

Semantic Tech & Business Conference Returns to San Francisco

Semantic Tech & Business Conference returns to San Francisco in June! Join us from June 3-7 for complete coverage of Big Data, Linked Data, Extreme Information Management, and Semantic Web. From breakthrough approaches to solving business problems to the big data implications of fast–evolving technologies, SemTechBiz provides you with an unparalleled interactive experience and delivers tangible business value. We're offering a special early rate when you register by February 17. Sign up now!