COMMENT Electronic health records are too complicated to emulate the success of NHS Direct, believes Peter Mitchell

The other day, on Radio 4's You and Yours , I heard yet another GP wheeled on to snipe at NHS Direct. How inconvenient it was, he said, that NHS Direct nurses were giving advice to patients that sometimes conflicted with what he himself told them (which, he implied, was always utterly reliable and derived from the very spring and fount of leading-edge clinical knowledge). How much more sensible and effective it would have been, he said, for the Department of Health just to give each GP a few hundred thousand pounds so that practices could set up their own helplines and appointment booking systems, instead of constantly interfering with honest doctors.

No doubt this suggestion rang hollow with the millions of people whose GP practices still do not trouble even to answer their phones within an acceptable time, or for that matter do not work a proper appointments booking system - and whose patients therefore regularly waste hours trying to get to see them for five minutes.

It is not surprising that the medical profession is still unhappy about NHS Direct. It was, after all, introduced partly to undermine their grip on the service, and thus try to get around some of the chronic faults in the provision of primary care.

What is most galling for the profession is that NHS Direct is working rather well at what it does - bearing in mind that it does not purport to do everything. It just does what it can, and refers the rest elsewhere - rather like a GP, in fact, with the difference that NHS Direct is not continually demanding more money while simultaneously refusing to implement DoH policies.

The moral for the DoH is that information technology is good at performing certain simple tasks that make an enormous difference to the provision of effective care at relatively low cost.

Furthermore, once these basic, simple services are running reliably and are well understood, they can then be extended step by step. This is what is happening in North Tyneside, whose NHS Direct service is being developed in various directions to solve related problems - for example, by putting the service into hospital A&E departments to triage people as they come in.

It is true that NHS Direct has been implemented quickly, but not so quickly that the project became unmanageable. It has managed, largely because it relies on relatively simple technology that is well understood from elsewhere.

Contrast NHS Direct's success with the NHS's other major IT project - the plan to build a lifetime electronic health record, or EHR, for all citizens.

The EHR programme was announced as a grand strategy - a seven-year plan to get from square one to a vision of the NHS that is so ambitious that it exists nowhere in the world today. Not only is it technically very hard, but it is the ultimate in project management challenges, particularly in an organisation like the NHS, which is so unresponsive to central direction - and where there are mechanisms (such as the idiotic procurement procedures imposed on local bodies) working to prevent it ever happening at all. After all, it has taken 18 months just to get to the point at which EHR demonstrator sites are to be allocated funding.

It is clear that the EHR project will fail to meet its targets. It is far too ambitious. It is quite possible that it will fail completely - that some of the central technical problems, such as online access to primary care systems, cannot be solved.

It is ironic that the Cabinet Office recently announced it is launching a review of failed IT projects in the public sector - including the NHS - in the hope of learning lessons.

Perhaps the review team should extend its remit to looking at present projects, too, because EHR is another NHS IT disaster just waiting to happen. Comparing it with NHS Direct may reveal the lesson the NHS really needs to learn: that in information technology, five-year plans just do not work.