Go slowly, but get useful results quickly. That may sound like an incompatible pair of objectives for a regional health data-transfer initiative, but it describes the under pinning of a statewide effort coordinated by the Massachusetts Health Data Consortium.
It's based on the premise that, unlike a lot of regional information initiatives under way, the interlinking of data systems is the last step, no t the first.
Putting useful data on that regional link requires a foundation of common assumptions about the kinds of data to be shared and the specific elements to be included or excluded in exchanges.
The way the Massachusetts proj ect sees it, the foundation must first be built into participants' own developing information networks. The linking can come later.
By then, the hookups could well be accomplished by merely connecting compatible networks instead of building a costly new network from scratch.
The approach acknowledges the immediate problems that must be solved quickly in the construction of information systems at hospitals, HMOs and other area healthcare players. At the same ti me, it resolves to avoid committing to any one technology or vendor to create a regional data-sharing network.
The project's 15 participants (See chart) will be investing millions of dollars in their own internal system s, bulking up for the clinical and cost-control data they need to survive.
"The high penetration of managed care in this region (40%) has generated a sense of urgency to demonstrate how providers and health plans can collaborate on delivering new approaches to monitor quality, access and cost," said Elliot Stone, executive director of the Waltham, Mass.-based data consortium.
That imposes both internal and regional pressures for more and better data, said Ri chard Shoup, vice president of information systems at Tufts Associated Health Plans. "The market realities are we have clinical systems that are evolving, and at some point we know we're going to have to be able to share informati on," he said.
As providers and health plans develop their individual initiatives, they'll all face the same decision sooner or later on data standards-the content and technical specifications of electronic messages.
If those decisio ns can be coordinated at the outset, the players will in effect share their developmental costs and avoid paying extra to retool their information for regional sharing, Shoup said.
Long and short range.
The goal of the regional proje ct, called Affiliated Health Information Networks of New England, is a "network of information networks." Using an innovation referred to as a "virtual data repository," the project will aim to link networks by means of a compu terized central road map and formula for access to authorized information throughout the region.
There won't be a central repository of information. Data will be accessible to certain users under specific conditions, but each public or private health-related system retains ownership of its own information, Stone said.
In the meantime, the area's participants will inch toward that goal during the next several years by identifying immediate business problems rel ated to information-systems planning, Stone said. The coalition has identified four such problems that can be solved within a year or less: a clinical data set for the emergency department; practices for privacy, confidentiality an d security; a standard way of sending messages on insurance eligibility; and an approach for unique provider identification.
Work groups on each of the four initial business problems are staffed by key medical and systems representa tives from participating organizations, and they're charged with producing "results that are useful to the design of their own health information networks," Stone said.
By drawing on the cumulative brainpower of people working on the same problems, and by getting commitment from all participants to forge and adhere to a consensus, the effort can show immediate payback on investments of time and money regardless of where the regional effort leads, Shoup said.
But the increments of payback also will build momentum for the larger goal. "We're putting the hooks in place now so that when the time comes, we'll be able to share information," he said.
In other regional effo rts around the country, organizers have soothed participants with assurances that their internal information plans would not be influenced by regional decisions, as long as they became "CHIN-ready."
Competitors said they wanted to preserve whatever strategic advantage their information-systems plan could give them and go about building that advantage using whatever technologies and structures suited them.
But the go-it-alone approach to computerized records as a competitive advantage is "a fallacy," said Patrick Mattingly, M.D., vice president for strategic planning and development at Harvard Pilgrim Health Care.
Because of the cost and complexity involved, electronic records capacit y won't be created on a sufficient scale until providers get together to create them jointly, said Mattingly. Then they'll compete on the ability to use the results effectively.
The purpose of the coalition is to "centralize the as pects that are noncompetitive" and act as "a change agent for inter-operability and connectivity among public and private systems," Stone said.
The regional players have committed to fund a budget of $250,000 a year for three yea rs to plan that mission and hire staff. In July 1995, the data consortium named Lee Barrett, chairman of a national standards committee for insurance transactions, as director of the Affiliated Health Information Networks project, and the first work groups were convened.
Besides the common clinical, administrative and financial data-sharing tasks, the project has drawn some conclusions about how technology and vendors fit into the picture.
Initiatives in Chicago, Cincinnati and elsewhere have focused on selection of vendors to build the technical foundation. But the Massachusetts effort is putting vendors and technology in the background for now.
"Information systems vendors have an important supporting role, but they are not directing the activity, as may be the case in other states-the CIOs here have not relinquished that responsibility," Stone said.
What they'd rather do is make the technology in dependent of vendors and their products. "Pick the standards first, then the vendors don't matter," Stone said. By making vendors comply with customer specifications instead of the other way around, the approach can support multi ple vendors and a variety of technical and software solutions, he said.
And the decision to wait until the last moment to commit to a technical plan will allow the project to harness breakthroughs that don't exist yet.
For example, Shoup said, the emphasis today on building separate telecommunications networks for regional data routing may prove costly and unnecessary if future developments improve the means to exploit existing links, such as the Internet.
"By the time we're ready to exchange information, it could be a brand new landscape," he said.