Welcome to Notes from the Underground Railroad

Here's an amazing story, from Trotter and Uhlman's Meaningful Use and Beyond:

“years before the healthcare profession as a whole was aware of the dangers of Vioxx, the VA discovered on its own that it was a dangerous medication. Data from the VA’s electronic healthcare record, VistA, had alerted the VA that something was amiss with Vioxx. The VA took steps to ensure that Vioxx was prescribed only with careful monitoring and only in special circumstances, a drug of last resort. By doing so, the VA saved thousands of lives.”

A government agency, using open source software, used information technology to save thousands of lives.

When we lose thousands of lives in an accident or in war, it becomes a national tragedy.  Saving lives, however, is a much more subtle process.  The people whose lives were saved have no idea of that fact.  Dissolving problems before they manifest is a tricky business; solving them after they hit gets all the attention.

The backstory to this remarkable discovery is amazing.  Far from the top-down bureaucracy, beltway bandits, and cozy dinners with lobbyists, that folks normally associate with Washington,  this discovery was actually a cascade of events trigger from the bottom of the organization chart.  Today, this system is known as the VA's VistA electronic health record system.  About half of the electronic medical record systems in operation in the US today use this system.

For me, the story begins in June 1978, when I was having dinner at Coffee Dan's restaurant in Loma Linda, Ca.  I was dining with George Timson, a Harvard graduate/music major and one of the best programmers I've ever met.  We were talking about ways of using minicomputers - an exciting new technology - for a decentralized computer system to handle hospital information systems.

I drew an onion diagram on a placemat, the center of which was an incredibly simple core - which today would be called a "virtual machine."  This was a collection of 19 commands, 22 functions, and one data type, all controlled by a single language.  The next layer of the onion used this virtual machine to describe a "roadmap to the database" - which today we would call "metadata" - or data about data.  When technology changed all we had to do was adapt the virtual machine to the new approach, and the outer layers wouldn't be affected.

Next came the programs to interpret the metadata - which we called File Manager - which supported the next layer - Patient file, User file, and other files relating to clinical care.  The next layer we called "packages" to handle application-specific needs such as pharmacy, laboratory, radiology, and the like.

I worked as a computer specialist at the Loma Linda, CA VA hospital, and was a part of a group of programmers hired by Ted O'Neill and Marty Johnson, who ran a group called the Computer Assisted Systems Staff in the VA Central Office, reporting to the Chief Medical Director.

There was another group in Central Office who wanted to have a centralized mainframe computer in Austin run the VA, and were very unhappy at the notion of decentralized minicomputers popping up around the hospitals.

This soon turned into bureaucratic warfare between the centralists and the decentralists.  Ted O'Neill got fired, as did several others of my peers.  I was demoted, lost a programmer assistant, and was not allowed to open a $100,000 computer sitting in the basement of our hospital.  (I was forbidden from sharing software with others, which I did anyway, calling it "commiting portability".)

The centralist's attempts to shut us down gave us a common enemy to fight against - and a need to show the world that we could design a decentralized hospital computer system.

We came up with a system in 1982 that we showed to Donald Custis, the Chief Medical Director.  He had been briefed that the field developers were not able to pull together a system; that their efforts would lead to "helter skelter" development, and that only the professional staff of the central mainframe group could pull off something so complex as a hospital information system.

He saw a working, integrated prototype of what we now call VistA.  Rather than the helter-skelter development, he saw an integrated, cohesive system with a common user interface - and doctors who were enthusiastic about using it.

He quipped, "It looks like we have an Underground Railroad here."

I took this as a wonderful name for our previously unnamed gang... and had membership cards printed up to hand out.  I began to hold Underground Railroad Banquets to recognize "Outstanding Engineering Achievement" and to offer  "Unlimited Free Passage" on the Underground Railroad.

How we progressed from an Underground Railroad to saving lives with by discovering problems with Vioxx is a long, involved story.  It is not just a history and nostalgia, but also a lesson for today and the future.

I hope to make Notes on the Underground Railroad a enjoyable romp through this history, as well as a vision of what the next 33 years of Health IT might bring.


MUMPS Underground & VA VistA

Great blog. The story of the MUMPS Underground movement that helped bring about the development and implementation of the VistA System at the U.S. Department of Veterans Affairs reads like a movie script. See a related OHN article on VistA & the Underground Railroad at http://www.openhealthnews.com/hotnews/vista-underground-railroad