Published: 27/05/2004, Volume II4, No. 5907 Page 22 23

HSJ notes that there are problems with clinical involvement in the national electronic patient record system currently under development (The Big Story, pages 10-11, 6 May). This is hardly a surprise, not least because the evidence to demonstrate improved outcomes from such records is not easy to find.

Of course, they can save paper and time for those regularly seeing a patient. But my understanding is that the project has ambitious plans to give a wide range of providers access to a patient's record. This is fraught with all kinds of practical problems, including whether a patient with a common name has changed their address.

It also raises questions about whether it would be cheaper and easier to extend existing alert systems so that a provider knows important health issues about a patient quickly, eg using cards and bracelets.

The evaluation of health outcomes achieved by electronic patient records is in its infancy. In a study limited by the scope of the system involved (BMJ, April 2004; 328: 875-878), researchers in the south of England found that, for example, 54 per cent of records on a usefulness scale running from one (no use) to five (very useful) were scored as one or two. This excluded the large percentage of records that were not found, though that was potentially due to coverage.

It is difficult to find any kind of solid research base on the outcomes achieved by rapid access to electronic medical records.

Perhaps this is another case where we should collect evidence rather than plan on a massive and very complex implementation.

Peter West Director York Health Economics Consortium Ltd