By Scott Shreeve

While on my VistA kick (here and here), I need to respond to several important errors of understanding in the recent press release hailed with a “Bravo!” from Fred Trotter. I also wanted to take the opportunity to mention a significantly broader and more meaningful opportunity that the open source community should be rallying around.

First, so people are clear – the Department of Defense does NOT currently use VistA. They haven’t since their 1988 decision to have SAIC fork the code. The only reason that VistA is mentioned as part of the DoD’s selection process is that their own physicians are clamoring to throw away the current system in favor of VistA. While the DoD is correct in identifying some of the weakness of VistA, they also appear to recognize many of its outstanding clinical attributes.

Comments from a July 21 letter from Principal Deputy Assistant Secretary of Defense for Health Affairs Stephen Jones seem to indicate a ray of hope for a VistA compromise: “There is a strong feeling here and at the VA that the best approach is a convergent evolution of the two systems. This approach optimizes the strengths of both systems while creating interoperability that will drive more universal information exchange.“

Second, AHLTA is the second billion dollar proprietary EHR that the DoD has commissioned and subsequently failed miserably with. When the DoD committed their original sin in 1988 with SAIC, they chose to disfigure VistA to the proprietary point of no return (and no interoperability).

The resulting CHCS system served the DoD in workman like fashion for a while, but ultimately succumbed to poor architecture, poorer design, and ultimately the voracious lobbyists who are fed by the even hungrier beltway bandits. To add insult to injury, the DoD once again made the decision to build an entirely new system from scratch, this time with Northrob Grubbin’. The new system, called AHLTA, has been roundly criticized, and continues to disappoint by all known subjective and objective metrics. It has not helped the DoD come closer to the goal of interoperability with the VA based on every GOA report that I could find (here, here, and here). Basically, it’s another several billion dollar disaster.

At the crossroads once again, the DoD is confounded with the choice of continuing to invest in AHLTA, invest in an off-the-rack private sector solution, or invest in the conversion of the VistA system to meet military needs.

Finally, as the DoD considers transition to an off-the-shelf proprietary product, it naturally raises questions about what the VA should do with VistA. By extension, it should also raise questions about what the government is doing with the red-headed step-child system developed by the Indian Health Service called RPMS.

The RPMS system is a gem in its own right, and had allowed the IHS to garner national attention for its clinical systems despite the remainder of the agency being exceptionally poorly run. Developers have labored in the shadows in doing some amazing work. Since they were so starved for federal funds, they were forced to stay within the same interoperability zip code as VistA (but still found a way to get creative with agency specific applications for their unique, at-risk populations). What emerged was an impressive system that has allowed the IHS to achieve VA-like improvements in clinical care despite an arguably sicker population (yes, it is possible!).

Considering these three systems should hearken us back to the halcyon golden-days of the Government Computer-Based Patient Record (GCPR) — the “one-true system to rule them all."

You might remember that this project was kicked off in 1998 following the Presidential call for VA and DOD to start developing a comprehensive, life-long medical record for each service member. As a first initiative, undertaken that same year, the GCPR project was envisioned to be an electronic interface that would allow physicians and other authorized users at VA and DOD health facilities to access data from any of the other agencies’ health information systems. Well, it turned out that the various agencies didn’t want to play nice, didn’t want to give up ground, and would rather play in their three separate sand boxes as opposed to collaborate together, pool their collective intelligence and collective budgets.

The concept or notion of leveraging the collective investment of all government-based health information systems is common sense. A common framework of shared services, with individual modular adaptation by department, and leveraging the collective weight of the agencies to drive home on standards of communication, lexicon, security, and purchasing. Instead of another 400 page white-paper describing the merits of describing a project, why not spend some mental synapses on the collective investment and collaborative framework from which a real run could be made at such a project. This is where an open source community could really have an impact.

To get sense of what I am describing, here is a money quote from one of the reports:

• The VA “has one integrated medical information system — the Veterans Health Information Systems and Technology Architecture, or VistA — which uses all electronic records and was developed in-house by VA clinicians and IT personnel. All VA medical facilities have access to all VistA information.”

• In contrast, the GAO said, “The DOD uses multiple legacy medical-information systems, all of which are commercial software products that are customized for specific uses. Until recently, those systems could not share information. In addition, not all of DOD’s medical information is electronic: Certain records are paper-based.”

Besides some fundamental inaccuracies, the press release is correct in making a plea to whomever will listen that the madness can be stopped by collaborating around a solution that leverages the investments by both organizations. Open-source development, collaboration, and governance processes, well established in other huge projects, could be a means whereby this unification vision becomes a reality.