IT could be the key to cutting waiting lists - but an out of date and ill-conceived strategy is not going to modernise anything, warns Dr Eamonn Butler

The biggest employer in europe is, unfortunately, too big to manage. Now it has embarked on the biggest IT procurement in Europe and that, unfortunately, will be too big to manage too.

Of course, the NHS was born in the postwar age, with its faith in central control. It grew up in the age of giantism, alongside Heath's super-ministries and super-councils. Its approach to IT is the same as its approach to healthcare: develop national plans, set national standards, deliver on the big scale.

But today technological change itself has made this whole approach quite out of date. The production-line days are gone: today's technology delivers us TV, telephony - even cars - precisely tailored to match our individual demands. Just go on to the web and order the one that is exactly right for you: size, colour - your choice.

And this customer-focused delivery is what people are starting to demand of our producer-focused state services. That is why they are having a hard time. That is why there is so much talk of "modernisation". But an ill-conceived IT strategy is not going to modernise anything.

These plans are too top-down to work. They talk about taking "greater control of the specification, procurement"and delivery of the IT agenda" with "strategic outsourcing" managed "at a national level" and "maximum use of national framework contracts". In other words, Richmond House (or at least the NHS Information Agency) is going to run the whole show. And a pretty rotten job they've made of it so far. When every other business in the country is computerised up to its eyeballs, the NHS still writes people's records on loose bits of paper and has to employ platoons of people to go down into the cellar to retrieve them (if they can find them).

If IT is so key, why has it taken so long to get to this stage? Kaiser Permanente, a US health plan covering 11 million (mostly blue-collar) workers, spends £750m a year on IT. That is three times what the NHS proposes to spend in this great push up to 2004. And it is one reason why waiting lists are unknown to its members.

I used to joke that we'd have made more progress if we'd simply given every nurse and every doctor access to a web-enabled PC and told them to sort it out for themselves. Pretty soon they'd have worked out through trial and error what was the most practical design for patient records, where to look for the best appointments systems, how to integrate with the pharmacists so that everyone was saved pointless prescription paperwork.

But it isn't a joke, given today's technology. Things can grow and evolve on the web. As medical treatments progress, so do patient records have to expand and adapt to keep step. If you'd been designing a patient record in 1948, you might have had a line for smallpox, but you wouldn't have had a section for cancer treatment. And the danger of today's planned top-down record design is that the pressure to "standardise" means it cannot adapt. If you leave it to the doctors and nurses at the patient end of the business, they will select the systems that they find most practical, and they will adapt and evolve them to keep abreast of those changing and local needs.

Stanley Kalms, the head of Dixons, once attributed his success to tearing up any memo containing the word "strategic". It is what happens at the level of the customer, or the patient, that is important. If we are to have a "strategic" IT plan, it would be best delivered through the PCTs, as independent and dominant purchasers who are close to patients. They could, in fact, drive a revolution in how healthcare is delivered, forcing other providers to work in new ways. They would be looking for best value, and greater clinical and customer-service benefit to what they spend, than the average Whitehall department, quango, or strategic authority.

What we should be thinking about is not how best to wire up our hospitals, but whether today's IT means we don't even need hospitals. Just how much care can be devolved to the local level, thanks to technology? Or even delivered in the patient's own home through tele-diagnostics and tele-medicine?

Could radiology and other diagnostics be faster, cheaper, and more accurate if images and data were wired straight up to specialist reader centres instead of each hospital having to maintain its own staff? Can emergency services be revolutionised by stationing vehicles all over town, using IT to track and direct them, rather than having just one depot - at the hospital, where patients want to end up, not to start from?

Thought through like this, IT could produce some very radical modernisation of working practices. It won't, though, because this is yet another over-centralised plan to make NHS delivery conform to some central conception. Why can't our ministers and civil servants learn to let go and allow those who are much nearer to the action to run things as they believe right for the local people who use them?