EA serves NIH's clinical diversity
- By Mary Mosquera
- Mar 02, 2004
Jones has two approaches to meeting the challenge of NIH's diverse IT needs: accommodate the IT requests of scientists in NIH's mission, but consolidate support systems.
The National Institutes of Health faces a mixed challenge in IT management.
It must meet the needs of a diverse group of scientists at clinical institutes studying AIDS, cancer, and blood, heart and many other diseases. But the agency also is working to consolidate systems and develop an enterprise architecture.
Congress funds each of the 27 clinical institutes individually, fostering the independence of scientists who are accustomed to making IT decisions to support their research.
To promote cooperation with the agency's enterprise architecture and systems consolidation, Jack Jones, acting deputy director of NIH's Center for IT, borrowed a lesson from the Internet Corporation for Assigned Names and Numbers. ICANN, based in Marina Del Rey, Calif., is a nonprofit group of technical, academic and business leaders that establishes Internet policy.
The group uses communication and consensus to make decisions, and brings together disparate stakeholders with common needs to build standards.
Jones has similarly brought together representatives of NIH institutes in communities of shared interests to determine the business architecture.
'I'm trying to understand how we might identify standards and overall design for information at NIH that accommodates all of the actual needs of that diverse population,' he said. Architecture is really a systems design problem to be solved with requirements that stem from the business processes used by the enterprise, he said.
'We've made enough progress now that we're about to start sharing it with others to try it out,' he said.
Essentially, the plan represents the executive level view of the significant processes and data classes across NIH. Jones' group has also created a taxonomy'classifications of data to simplify organization'of all that needs to be addressed in the architecture.
Jones assembled domain teams from across NIH to examine the technology and standards needs of areas that were about to undergo significant consolidation, such as e-mail systems and wireless networks. Domain teams are now studying enterprise systems monitoring and soon will tackle selected application integration.
NIH has already experienced benefits from developing its enterprise architecture. During e-mail consolidation, NIH also combined some directory services, and Jones' team installed a new log-in function. Now NIH has about 10 different systems that share some common authentication infrastructure.
By spring or early summer, Jones will begin to expand NIH's network operations center to monitor three enterprise systems. The systems will have automated sensors in place to measure whether they are working correctly.
The next step is for Jones and other IT officials to determine the NIH business architecture.
Jones has a dual approach to meeting the challenge of NIH's diverse IT needs: accommodate the IT requests of scientists in NIH's mission, but consolidate support systems for functions such as financial management.
'Often when people talk about consolidation, they talk about bringing together the things that all of us do and ignoring the things that only a few of us do,' he said. 'You have to think about what happens to the things that you need to do when you are the only one. Are you now left high and dry with no way to accomplish those?'
Users tend to resist consolidation when they think the change will hamper them in doing their jobs. 'What is unique here is that we're trying to create an architecture in which, even if the central system doesn't do it, we provide the interfaces that will allow the individual to still accomplish those jobs in some way,' Jones said.
NIH also has adopted ICANN's system for requesting comments, which includes writing documents and publicizing them. 'That is a way for concerns that haven't risen to the enterprise level or can't be addressed with central funding to be discussed and for consensus standards to be built in areas we can't get to yet,' he said.
Mary Mosquera is a reporter for Federal Computer Week.