Hello Richard,
Actually, that's what the VA's central computing authority wanted to do from the beginning, but as a 7 or 8 year study by the NIH demonstrated, that doesn't work for health informatics, where every hospital works differently. One of the heads of that study showed up at the VA, got clued into the politics, spent a year networking with the best and the brightest face to face, and called a secret conclave in the Midwest. Using the study recommendations as a template, they agreed on an architecture and divvied up the work in one week, and 18 months later they had stitched it together and it ran.
The problem with centrally run is just that -- centrally run, by humans with the typical political instincts (me on top -- you do what I say) of animals who only split off from a common ancestor with the chimpanzees maybe 1.5 million years ago, who simply don't get that their job is either to teach the tech or doctor on the front lines, the point of care, how to write it themselves, or to serve them faithfully in making exactly the tool they want. A, that was anathema to the power-hungry, and B, the central model simply would take too much resources to do that, even if it were done in MUMPS. Done in the commercial software available on the market, it simply would not happen.
So, instead, the guerrilla movement, which ultimately was awarded the name, The Underground Railroad, taught its own, who iterated improvements continually. Now, the VA is rated the highest large hospital system in the US on something like 26 measures of quality of care. Crowd-sourced software in an environment which is controlled for the benefit of, and to serve, the people at the point of care, VistA EHR is the most significant pioneering, open-source, agile project ever, to the best of my knowledge. While the amount of content in Wikipedia dwarfs it, the procedural complexity and manageability together of VistA EHR dwarfs anything else I know of. It is also far faster, more stable, more resilient, and runs on less resources than anything near its size and complexity.
Actually, VistA EHR is a hybrid of the two models, skewed way to the left. There is a layer of administrators and programmers who support the techs and doctors in writing or modifying code, and there is a core team which dictates what gets in and what stays out of a build. But they don't do it with the typical psychology of bosses; they do it for the greater good; the good of the community. It is actually a communitarian ethic, through and through.
So that's my version of the case at that level! For a relatively rigorous version, you might want to read Best Care Anywhere: Why VA Health Care is Better than Yours, Second Edition, by Phillip Longman, 2010, Polypoint Press, <
www.P3books.com>. It has lots of references.
Regards,
Mark