Pragmatic Open Source: An Open-source System for the VA Is the Next Logical Step

A committee helping the Department of Veterans Affairs figure out how to modernize its core system unanimously recommends open source. Makes sense.

Arguably the gold standard of electric health records management, the federal Department of Veterans Affairs could be "re-engineered" as an open-source project.

That was the unanimous recommendation of the Industry Advisory Council, which was asked to study and suggest ways to modernize and simplify the VA's VistA (VA Health Information Systems and Technology Architecture), which serves about 8 million veterans through 153 medical centers and 768 outpatient clinics across the U.S.

VistA has been continually developed and upgraded over the past 32 years, but if it's to continue being effective, it needs an overhaul. Updates on top of updates have rendered VistA difficult to maintain. So the VA decided to seek recommendations for an overhaul before it needed fixing.

The IAC committee unanimously agreed on the report's key recommendations:

  • VA should commit to and announce as a matter of strategic policy a plan to move to an open source, open standards model for the reengineering of the next generation of VistA.
  • VA should contract with one or more Federally Funded Research and Development Corporations (FFRDCs) to:
    • Provide a detailed set of technical recommendations for the development of a VistA 2.0 Open Source Ecosystem;
    • Assemble a fully functioning application development environment for use in the VistA 2.0 Open Source Ecosystem;
    • Develop one or more demonstration applications using the Open Source Ecosystem and the application development environment; and
    • Provide an appropriate Concept of Operations which includes a business model, charter, bylaws, operating principles and organizational blueprint for an independent, not-for-profit Open Source Foundation to manage, operate and maintain the VistA 2.0 Ecosystem.
  • The current VistA should be placed on an aggressive program of stabilization with limited tactical upgrades and enhancements driven by patient safety and other mandated requirements.

The idea is to have the system appear to function much as it already does, but with streamlined code and the ability to make updates to it more easily. This would allow greater ability of VA employees, healthcare professionals and others involved in the VA's services, such as large contractors and healthcare software vendors to contribute to the system's improvement.

The council made several suggestions on how to find the right people to launch this re-engineering, such as the "Toucan" (two can) approach, which is how much of VistA was developed in the first place: two people, one highly technical and the other end-user-oriented (possibly a user himself) working as a team; or "Structured Open Source," where the VA would "aggressively participate and even lead the open source community" in developing the new systems, applications and code.

In the end, creating a modular system, where bits and pieces could be added or removed without affecting the overall funcationality, working in tandem with a greater open source community, was pointed to as the best model.

Fact is, right now the VistA system works great and is comprehensive, everyone agreed. But its roots make it too hard to innovate on the system and deliver new functions or integrate new technology in a timely manner. This issue was addressed, somewhat in the VA's open government plan released recently, which got "middle ground" ratings on the audit by

A new VistA system would have a basic kernel that would be well-documented so in future years, no one would have to depend on one or two people to make significant changes to the guts.

If that's not a solid argument for moving to an open source infrastructure, then there just isn't one.