Saturday, October 18, 2008

Best Care Anywhere, part 4 - the Hard Hats

Last time I finished recounting the "where we are today" stats, about how remarkably the VA has been outperforming other systems on measures of care quality and safety, all while maintaining costs. As I've continued reading I've been quite surprised at how they got there. It's not by any process anyone would design, and yet the methods have tremendous implications.

In the post-Vietnam years the VA health system was in lockjaw. The combination of politics and change-resistant bureaucrats made improvement all but impossible, just as predicted by the observers in the previous installment who felt any government operation was doomed to end up this way.

But for years, way under the radar, operating in basements (literally) and often going directly against orders from upstairs, a loose federation of geeks and medicos were hand-coding computer systems, largely out of self-defense - they were desperate to find ways to handle their impossible workload. They wrote programs that automated admissions and discharge, handled prescriptions electronically, maintained patient records to sniff out potential problems.

They wrote it all in open source: nobody owned the code, and anyone could check it for accuracy or contribute improvements. (For those who aren't familiar with the idea of Open Source software, which anyone can edit, try this: Wikipedia is an open source encyclopedia: anyone who knows how can go in and edit any page they want.) (Of course, someone else might edit it back. Such communities find ways of being civilized, and progress usually happens.)

Unlike today's sickening morass of systems that won't talk to each other, these people wrote programs specifically with the intent that the data should be shareable. Most important, since the programs were written solely to help the users (e.g. pharmacists), the only features that got programmed were ones that the users wanted, and the features worked the way the users wanted.

Mind you, this was before PCs. And to keep their work secret, these people, who came to be known as the Hard Hats, couldn't even write their software on computers - the requisitions would have been too conspicuous. So they often wrote their lines of code on primitive 1970s word processors, so there was no error-checking or system assistance at all. They'd type in the code and then carry the disks to a computer somewhere else, and only then find out if there was a typo or a logic error. (Writing software is much much easier today, but these guys did it blind.)

Management, the "high priests" of computers in those days, hated it. One of the Hard Hat leaders had his DEC (Digital Equipment Corp) computer piled with paper and set on fire; his car engine was rebuilt several times because someone repeatedly put sand or salt in the tank. Computers found to be used for this work were confiscated.

Meanwhile, in a great irony, the people who were squelching this competence-centered programming were incapable of getting anything done themselves, so the gulf between what worked and what was authorized grew wider and wider. In 1979 the VA's computer office removed more systems than it installed. (I had no idea this was going on!)

In 1981 it came to a head, when a broad rebellion arose, finally getting the attention of a top VA manager. Long story short, the situation was turned on its head: hundreds of high priests were fired and the Hard Hats were empowered. With the restraints removed, all the different disconnected programs the Hard Hats had written could be brought together into one big system - and the book says it took less than a week.

Most astounding of all, that combined system came to be called VistA - and it's still in use today. Where's the software anyone else was using in 1981?

What do we see here? I'll start, but I'd really like to hear what you see.

First: great systems are built specifically and only for the people who will use them. They're created by asking the users what they need, not by sitting in a conference room trying to anticipate what the user might like.

And the users are brought in early, to see the earliest prototype, which gets modified to work the way the user wants.

Second, extreme need can be a powerful creative force as resources become so overtaxed that people will do anything to relieve the pressure. If you look at how many boomers are about to enter the high-consumption phase of their healthcare "careers," you can read the handwriting on the wall.

Third, the advent of microcomputers and then PCs meant that power (computing power in this case, and power to create) got into the hands of the people, who could then do anything they wanted with that power. Look what happens when access to tools is unlocked, instead of being under centralized control.

It's power to the people, isn't it?

Here's something to think about. Today, with blogs, user forums, and all those other "Web 2.0" tools, we all have a shot at using unprecedented tools for finding, sharing, and publishing information. (Look what I'm doing right here, sharing what I see, and look what you're doing by reading it. This blog costs me nothing - you too.)

What else do you see in this story so far? What raises your eyebrows? What will you tell friends?

No comments:

Post a Comment

Your comments will be posted after they are reviewed by the moderator.