The health service is not alone in its ability to throw away millions of pounds - even tens of millions - on information technology projects simply by refusing to learn from its mistakes.

Remember the Passport Agency fiasco last year, when many holidaymakers failed to receive passports in time to travel? Officially, this was due to problems and delays with the agency's£120m computer projec t . It is less wel l known that 10 years earlier passport offices experienced a similar IT disaster, with a backlog of 500,000 unprocessed passport applications, after making precisely the same mistakes.

To try to avoid problems during a procurement, one trust visited a London teaching hospital to see how contracting worked on its new computer system, and received a fabulous review. The trust bought and installed the system, but ran into serious difficulties. It rang the reference site for help, but was amazed to hear that the module had been abandoned. Apparently the system had been beset with problems from day one, but the users did not like to say so. Now two organisations were locked into the nightmare instead of one.

The way the public sector manages its failures is a major reason why the NHS has been unable to benefit from IT in the way that the banking, insurance, and travel sectors have.

It is human nature to hide mistakes, but in NHS computing this seems to have become official policy. The NHS Executive prevented staff at its coding and classification centre issuing warnings about the Read coding system by invoking the Official Secrets Act. Many will have had access problems on NHS net this year, but the scale of the problem with this key infrastructure system has been obscured.

As well as hiding mistakes, the Executive has in the past provided disinformation. Despite an early draft of a National Audit Office report which was hyper-critical of HISS - the hospital information support system project - the information management group briefed a health minister that 'the vast majority of systems implemented have been installed successfully'.

Why does the NHS actively hide its mistakes? There are three main reasons: emotional equity, a blame culture and the internal market.

Emotional equity occurs when those close to a project become so personally involved that they cannot see its failings, or bear to see it criticised in any way.Team members cling to small victories and persuade others of their success, regardless of the facts.This theme repeats itself again and again in reports from the likes of the Audit Commission and the National Audit Office.

The point is beautifully illustrated by research into the NHS's£300m resource management project by Brunel University. The findings showed that the project had few benefits, and that 'these were not commensurate with the considerable efforts put in by the sites'. It went on:

'The public presentation of work at the sites was often out of step with (far ahead of ) actual progress.'

There was 'little evidence that the new arrangements had increased the efficiency or effectiveness of patient care'.

It continued : 'The experience of RM suggests that innovations that are inefficient and ineffective can be extended by determined stakeholders. The endorsement of RM appears to have created a bandwagon effect - and a policy which reduces efficiency and effectiveness has now been implemented across the NHS.'

The second prime cause of secrecy is the blame culture - or the NHS's difficulty in forgiving mistakes. This is linked to the limited rewards for success, and is clearly evident in the stream of letters that can come out of NHS organisations during projects.

Writing letters that cover the author in case of problems has become an art form. If anything should go wrong, there will always be a letter on file somewhere which covers the manager concerned. Self-preservation in the event of failure can become more important than preventing failure. On a national basis, this can be seen in the 'spend a pound to save a penny' mentality that typifies central control schemes such as POISE (procurement of information systems effectively).

Finally, there's the internal market.

The introduction of competition effectively put an end to the spirit of inter-organisational co-operation for which the NHS was renowned.

Organisations that had been used to sharing information, experience and even resources freely became far more circumspect. The increasing involvement of consultants and suppliers who wanted to protect commercial 'secrets' exacerbated the problem.

But there are positive signs. With the demise of fundholding, and the introduction of initiatives such as local implementation strategies, health organisations are working with one another once again. There is also a move towards true supplier-trust partnerships in IT projects.

This partnership approach is essential if we are to unlock the full potential of IT in the NHS. The breakneck speed of change in computing means that best practice can become worst practice overnight.

Computerised prescribing and multidisciplinary care planning are both good cases in point. Despite the fact that both are relatively simple applications for computer sales people to make look glitzy in a product demonstration, they are both notoriously difficult to implement operationally.

In order to progress, we need to pool our experience of what worked, what didn't work, and what still doesn't work. We have to share knowledge to keep up, and - whether trust or supplier - we need to be able to do so without fear of being pilloried.

Openness is both our single most important weapon against serial failure, and a critical factor in ensuring that the NHS gets the best possible IT that it can afford.

See IT Update, centre pages.

Markus Bolton is chief executive of a healthcare computer systems supplier.