Yes Minister was just a TV programme, wasn't it? Sir Humphrey and his selfperpetuating, relentlessly expanding army of mandarins do not exist, do they? We can rely on our public servants to get IT policy in the NHS right, without charging us too much for the privilege, can't we?

When Labour was in opposition, some policy work was done on the future of the NHS Executive and its information management group. It was generally thought that the IMG was too large, too powerful, and too centralising and should be broken up into smaller, cheaper units, one of which would help set central standards for IT systems, while another helped provide consultancy support to trusts.

Labour's policy was generally well received by clinicians, managers and IT providers.

When Labour took office, the IMG cost£30m and employed about 400 staff. Plenty of scope for savings there, you might think.

But after a year, what did we find? The new NHS Information Authority had grown and employed 500 people, with an annual cost of£45m.Meanwhile, a separate information policy unit had become part of the Department of Health. So there were two bureaucracies where only one existed before. And just one of the two now cost more than the whole of the old system.

So it continued. In the financial year 2000-01, the NHSIA had a budget of£77m. Now it is well over£100m and the personnel requirement is over 700.

As Sir Humphrey would recommend, this cheerful empire-building took place behind an obscuring barrage of rhetoric about rationalisation, belt-tightening and streamlining.

There was constant publicity about the strategy of streamlining the authority from 23 operational locations to 11. But as usual, Sir Humphrey's words distracted ministers from his sleight of hand. There were only ever seven main centres, and a number of insignificant small offices. These small offices were easily closed.

Meanwhile, the six main centres grew dramatically. The essential new corporate HQ, Aqueous 2, is particularly splendid.Yes minister, it looks smaller, but ha ha, really It is bigger. . .

The principle behind the consolidation of the IMG was probably correct and It is interesting to see a government take its election promises seriously. It is also clear we need an effective unit to manage essential infrastructure projects.

We need the NHS number, the Exeter system, connected NHS professionals and a centrally managed electronic drugs database. Progress is being made in many areas, but it is not obvious that the use of resources is either effective or efficient.

The first two years of the new structure were spent in a stand-up row about responsibilities between the NHSIA and the IPU.

The IPU was supposed to focus on policy issues and on monitoring the NHSIA. But its staff came from operational backgrounds, running central IT projects. The NHSIA was meant to focus on delivery. But it complained that the IPU was always mean about staff and funds.

While this civil war raged, the danger was that responsibility fell through the yawning gap between the two bodies. Faced with a problem, the NHSIA could say it was because the policy had not been decided by the IPU, and the IPU could say that the NHSIA had failed on delivery. Jonathan Swift wrote about Little and Big-Endians who went to war over which end of a boiled egg should be cracked first.He would have found NHS IT policy a rich source for satire.

Meanwhile, there was extensive duplication of effort at conferences, meetings and exhibitions. Everyone else in the NHS was by now totally confused about which body was doing what. The organisations themselves seemed confused. I went to one conference where an IPU speaker announced that EDIFACT would be withdrawn an hour before an NHSIA speaker launched the new EDIFACT results service.

Ministers had hoped both the NHSIA and the IPU would help deliver national standards, negotiate better deals on national infrastructure projects, and provide people with IT skills.

Little of this is being done.

Both bodies have had to rely extensively on contractors and consultants, and are only now beginning to address project costing in a sensible way. Neither body can deliver effectively on national projects. They are hampered by trying to build a national consensus before they start anything. By the time they succeed in this, the rules of the game have usually changed. Both bodies are risk-averse, and projects have too many control loops built into them to proceed quickly.Meanwhile, government NHS IT targets are not being met.

Oh dear. Sir Humphrey triumphs again. Time for a new policy document, perhaps? l Markus Bolton is chief executive of a healthcare computer systems supplier.