Published: 21/03/2002, Volume II2, No. 5797 Page 13

A great deal of time and energy has been spent on providing an IT infrastructure that can support some of the concepts of an electronic patient or health record.

But one dimension has yet to be fully addressed.This is the way clinicians work and how, in particular, they record clinical information.

Medical students are taught to start with a blank piece of paper, ignore any previous information about the patient they are dealing with and record the information they collect - under headings but using free text. In this sense, the clinical record is a means of recalling individual patients, not a systematic recording of pre-defined clinical information.This lack of systematic recording of decision-making information is by no means limited to healthcare.Most professionals handle most of the information they need to make decisions in their heads.

To make this point, I often ask managers: how easy would it be to do your job if you had to predefine your data items, develop a systematic way of recording them and ensure data entry is complete and accurate?

Not easy.But this challenge to healthcare professionals is contained in the Information for Health clinical governance and revalidation agendas.Some initiatives do address this problem.

Local programmes, including electronic record development and implementation work, have resulted in working examples.

The regional learning networks, as they evolve, will allow healthcare professionals to understand, influence and develop local systems.The collaborative programmes will help to facilitate the introduction of datasets and possibly influence clinical practice to allow the collection of the data items on the datasets.

However, there is only limited understanding of this agenda at trust board level, with little or no certainty as to how to address it.

The EHR is clearly going to be fed from several sources, whatever its final architecture.

If several information views of the patient are presented - for example, 'backache' from the GP, 'PID' from the junior doctor discharge summary, 'prolapsed intervertebral disc' from the orthopaedic surgeon and 'pelvic inflammatory disease' from the gynaecologist - which of these 'views'of the patient is 'correct'?

Not only must data be entered in the correct field but also there will have to be accuracy, as well as agreement, across the clinical professions and disciplines about what is acceptable.

Clinical practice will have to undergo a revolution if the very laudable goals of achieving a universal clinical record, available in real time and at the point of care, are to be achieved.

I believe it can be, but equal - or possibly more - energy and resources will be required than have already been invested in infrastructure. Failure to do this would render the earlier investment worthless.

Bruce Hutchcroft is clinical relations consultant to the NHS Information Authority and consultant physician at Sheffield Teaching Hospitals trust.