Biggest IT Disaster Ever?

lizzyben writes, "Baseline has a major story about a major IT disaster in the UK: 'In 2002, the English government embarked on a $12 billion effort to transform its health-care system with information technology. But the country's oversight agency now puts that figure at $24 billion, and two Members of Parliament say the project is "sleepwalking toward disaster"... In scale, the project... (NPfIT) is overwhelming. Initiated in 2002, the NPfIT is a 10-year project to build new computer systems that would connect more than 100,000 doctors, 380,000 nurses and 50,000 other health-care professionals; allow for the electronic storage and retrieval of patient medical records; permit patients to set up appointments via their computers; and let doctors electronically transmit prescriptions to local pharmacies.'" An Infoworld article from earlier this year sketches some of the all-time greatest IT meltdowns.

Comments

depends on how you look at it

[User 956]

'In 2002, the English government embarked on a $12 billion effort to transform its health-care system with information technology. But the country's oversight agency now puts that figure at $24 billion

I imagine if you're the company getting paid the $24 billion, the project is a tremendous success.

Overarching

[AKAImBatman]

The word "overarching" comes to mind whenever I hear about projects like this. If there's anything my years in software and systems has taught me, it's that starting with a monolithic design will mean monolithic failure. You invariably end up with too many cooks, all working on Lord knows what, accomplishing a very expensive nothing. There's just too much coordination to maintain any semblence of progress.

In fact, the most successful large scale projects always seem to be grown out of combinations of smaller architectures rather than a single massive architecture. Look at the Internet for an example. The protocol was architected. The routing design was architected. The information delivery systems were architected. The network itself? Grown with tender loving care, and Lots'o'peering agreements.

If you want to solve an issue like modernizing Hospital IT, start small and work your way up. Design each technology independently, but not monolithically. Keep an eye toward standards rather than specific implementations. (Standards will allow you to plug in a few competing implementations, giving you "best of breed" options.) Then use those technologies to build out a few test sites. Work out the kinks, then start deploying at a few more sites. Keep doing that, and the economics of scale will begin to take hold. (i.e. The more you do of something, the less expensive it gets to do it.) With any luck, the project will get done within a reasonable budget and timeline.

In late September, Accenture, the global management and technology consultancy

Never mind what I just said. There's your answer right there. :P