Shortly after he announced the launch of the biggest civilian computer project in the world, Tony Blair told journalists: "The possibilities are enormous if we can get this right."
He should have added that the pitfalls were just as big.
Nine years and a new Government later, Mr Blair's vision for a joined-up, computerised NHS now looks recklessly optimistic.
But how did things go so badly wrong?
Amid the twists, turns and travails of the NHS IT programme over the years, several broad problems have emerged.
The first should have set alarm bells ringing from the off. Under the original proposal, all parts of the NHS would be connected by one single computer system, procured and maintained centrally and allowing total integration of information.
If you lived in London and fell under a bus in Newcastle, so the argument went, all your medical records would be instantly available at the touch of a button to the doctors treating you.
But when the teams putting the system together tried to work out what it needed to be able to do, they found it far more complicated than they originally thought.
Some hospitals were working with 30 computer systems and soon discovered that merging them all into a "one size fits all" IT scheme was hugely challenging – if not impossible – to achieve.
The Government responded by trying to increase the size and the scope of the system to take into account all the different requirements and concerns, and that led to a second problem.
Under EU law, government procurement contracts have to be enormously specific in outlining the scale and scope of the work to be done. The idea behind this is to encourage fair and open competition – but in the case of such a mammoth IT contract it was a huge hostage to fortune.
When the Department of Health changed the specifications of the programme – which it has done many times since 2002 – it incurred extra costs to the taxpayer.
And when it reorganised NHS structures, that created bigger bills and bigger technical challenges to what was already a hugely ambitious project.
So where to go from here?
Curiously, the delays in implementing the IT project may have been its saving grace.
In 2002 it was believed that a single IT system speaking the same "language" was the only way to integrate the NHS.
Now, due to technological advances it is possible to have many different systems which can still speak to one another.
And therein may lie the solution. As Tony Blair would put it, the possibilities are enormous.