Last year produced a flurry of activity toward promoting national standards for health information. Although the announcements, reports, and proposals flew thick and fast, they also illustrated there's still a devil around, still waiting for the details.
Those involved in the standards efforts as part of the larger national health information network believe the bulk of direction, funding and leadership must come from the top down — meaning that the Centers for Medicare and Medicaid Services must use their clout as the largest buyer of healthcare services in the country to dictate the direction of the standards effort. Others believe that the most profound advances will come from an organic growth of local and regional "circles of trust." Expanding these circles, however, will require not only widely accepted healthcare-specific standards, but also more maturity among the specifications underpinning more generic service-oriented architectures and federated identity technologies.
"For the first time in history, you have government, academia, payers, providers —everybody — really aligned with this principle that electronic medical records that are interoperable will reduce costs and improve quality," Halamka says. "You have not had much more than just talk in the past, but that talk has been translated into action. Action means there is government funding."
In October, HITSP received $3.3 million from the Department of Health and Human Services (HHS) to harmonize standards for specific uses; Halamka hopes the panel will have some recommendations ready some time this summer. A month later, HHS awarded $18.6 million to four regional consortia, led by Accenture, NYC; Computer Sciences Corp., El Segundo, Calif.; IBM Corp., Armonk, N.Y.; and Northrop Grumman Corp., Los Angeles to design health information network prototypes. HHS officials, in announcing the consortia awards, said the contracts "complete the foundation for an interoperable, standards-based network for the secure exchange of healthcare information."
However, some say the funding is a fraction of what is truly needed.
"The amount of traction the consortia get will be very limited," says George Hill, vice president and senior analyst at Leerink Swann & Co., a Boston-based healthcare investment firm. "To put this in perspective, the federal government is awarding $18.6 million to the four consortia. Kansas University Medical Center is going to spend $40 million to wire itself. The amount of resources they're getting is being spread too thin."
Don Detmer, M.D., president and CEO of the American Medical Informatics Association, Bethesda, Md., concurs with Hill. "What's clearly not getting through is the fact that if you're paying half the bills, you have to be willing to pay something to get this transition funded," Detmer says. "We haven't really realized this is really the equivalent of the Interstate highway system, and people act as though it's something that should be simpler than that, and it's not."
Funding is not the only issue. William Stead, M.D., director of the Informatics Center at Vanderbilt University, Nashville, Tenn., says the overarching standards effort needs a "big picture" architect who will oversee what individual groups are doing and suggest ways to coordinate their efforts.
"That's what's missing," Stead says. "We developed that as a model, and we frankly think there need to be multiple such models and they need to be used for proactive management and course correction."
Hints of progress:
Claims attachments One concrete example of fruitful inter-organizational collaboration is the work of the Accredited Standards Committee X12N and Health Level 7 (HL7) in creating the proposed standard for claims attachments under the Health Insurance Portability and Accountability Act (HIPAA). The proposal calls for containing HL7's Clinical Document Architecture within the X12N standard documents for requesting or supplying additional information supporting a claim.
On Sept. 23, HHS issued its Notice of Proposed Rulemaking for claims attachments, stating the complementary niches of the two specifications, plus the Logical Observation Identifiers Names and Codes (LOINC)—owned by the Regenstrief Institute, Indianapolis—yielded a robust combination blending the best of each.
This multi-organizational approach is encouraging to leading informaticians, but also gives them pause. The intricacies of designing interoperable administrative and clinical systems containing elements of such discrete code sets is still a ticklish area, notes Robert Dolin, M.D., a member of the HL7 board of directors and the editorial board for SNOMED (Systematized Nomenclature of Medicine), operated by the College of American Pathologists (CAP), Northfield, Ill. Owners of various terminologies and standards groups will have to devise a collaborative approach, he says.
"Just as we want HL7 and ASTM working together so we don't have vendors having to decide whether to use one over the other, likewise I think we want to have CAP and Regenstreif working together so we don't have to choose. Lab LOINC integrates nicely with SNOMED."
But clinical LOINC codes overlap occasionally with SNOMED, Dolin says, adding complexity in designing interoperable systems. It would help, he suggests, to have higher-level efforts such as those sponsored by HL7 replace one-off mapping attempts by vendors, providers or payers.
"I don't think you want everyone using SNOMED in their clinical applications to have to build their own mapping to LOINC code," he says. "I think this is a really positive step."
Many say an incremental approach at both the national and regional levels might result in small but significant advances both technologically and organizationally. As providers and payers begin exchanging claims attachment data over existing connections, they will be able to take advantage of well-established technologies such as the HL7 v2 family. However, as organizations expand their circles of communication, the logical choice for data exchange will be HL7 v3, with its Reference Information Model and comprehensive XML architecture, explains incoming HL7 Board Chairman Chuck Meyer. HITSP's Halamka says stakeholders' willingness to collaborate should hearten those who have been waiting for real progress on standardized interoperability, but that the revolution will be evolutionary in nature and far broader in scope than in just healthcare-specific technologies, and will have to include wider IT standardization efforts such as those around federated identity and the Web services stack. "We're building a network of networks based on regional, personal relationships that will ultimately result in a national architecture, not starting from the top down," he says. "There's no question we're breaking trail. We know perfection is the enemy of the good. Let's see if we can just come up with something that works and not have to wait for the ultimate WS [Web services] stack that won't be here until 2010."
Greg Goth is a freelance writer in Oakville, Conn.
