Those who digitise in haste will repent at expensive leisure, warns Dr Grant Kelly

Have yourself some fun: ask an interested person what an electronic patient record is, where it is, and how it is used. Sit back and enjoy the mounting confusion until kindness forces you to confess that you don't know either.

You have to be brave to question the dogma that the future NHS will be an e-NHS, but if you do, you are making this point: we don't really know what we want. Our problem is that we are digitising an NHS which has no corporate sense of history and, by turn, no corporate sense of future either. While IT schemes for industry have shareholders and directors focusing them on commercial development, the NHS has no such roadmap. Although we all know who signs the cheques (thank you, Gordon), nobody knows who the real customer is, so we don't know who to design the e-NHS for. Without this basic requirement, it's difficult to know which way to wire it up and what to wire first. Add to the mix that for middle management, the NHS is not a reward culture but a blame culture, its previous IT disasters, its inability to write competent IT contracts - and you can see that the planning for this monster e-task is a risky affair.

Being optimistic, it's obvious: can anyone seriously believe that in 20 years time, doctors and nurses should still be writing patient histories and results on self-losing bits of paper? Will they still be scouring books for evidence, rather than being prompted electronically about generic evidence and then particular points relevant to that patient? Will they be able to access patient information when and where it's needed? And what about the administration, who need to know (anonymised) information to provide for the service?

We just have to get there, but it's in defining "there" that things fall apart. Our proposed e-NHS is a more amorphous and uncertain body than anything that's gone before. For this proposed complex monster, we need a wholly different provision of electronic information, but neither the information users nor the system providers know what they want, at least beyond simple profit/cost. What we do know is that this task hasn't been done before by anybody else, that UK plc hasn't thought it through, and that we are in danger of proving the old adage that if you add electronics to a poor work process you will make it worse, not better. Keeping this in mind as we develop an e-NHS will not be easy when a "true" e-NHS needs a 10 or 20-year development timescale while being tied to a political control cycle of three to four years. In practice, this time mismatch is crippling.

There is a similar disparity in aims. Politicians and higher management hope that the e-NHS will cut fraud, increase efficiency, reduce wastage, cut waiting times, improve the issue of guidance to the service and provide better information returns to the centre.

At the other end, clinical staff of all persuasions don't know this. They want an e-NHS which doesn't interfere with their work, and just might help them. Although as taxpayers they disapprove of fraud, they are far too busy to be interested in it; they are already working at efficiency rates beyond any other country and know that further efficiency gains will come through changed working practices rather than IT. They know that reducing wastage is a function of forethought rather than afterthought; they know that cutting waiting times is a function of coherent planning of staff, premises, education and finance, rather than a function of IT "solutions". They know that e-booking creates work for overloaded primary care whilst not reducing the overall patient wait or improving patient care. They want proof that e-prescribing shows real patient gains which are cost-effective when independently audited; they know that e-mail is a curse in clinical practice, not a blessing. They know that internet technology supports multiple organisations without common aims or management and so is ideally suited to the NHS - but they also know that the NHS doesn't realise that. In short, for the most part, they remain to be convinced that the proposed e-NHS will help them do a better job.

What we need to do is agree the perceived gains of an e-NHS and then the path, the stages, and the long time it will take to get there. This won't be easy but it is clear that the e-NHS is not a technical problem requiring imposed "solutions", but very much one of management understanding work processes and problems. We can then identify where information management can help, gain agreement and move forward as the workforce comes on board. Those who digitise in haste and for the wrong reasons, repent at expensive leisure.