A delicate operation

There will be no more difficult proving ground for public sector IT than the NHS. Gordon Brown's gamble on National Insurance-funded health service reforms will rest fundamentally on technology.

The Treasury's recent review of the health service suggests that spending will double next year to £2.2bn.

The initial portents are not promising, however. NHS IT is certain to become a political football before the next election.

Internal dissent is also strong. Unions and other organisations representing medical staff have already made clear that they believe money should be spent on frontline staff rather than IT.

The single biggest benefit of information technology, of course, should be to improve the efficiency of that front line and to take on the burden of much vital but time-consuming bureaucracy. The argument would be more convincing, however, if the NHS did not have a history of failed IT implementations, electronic patient records being the latest in the series (see below).

Plenty of people recall the collapse 20 years ago of the Wessex Regional Health Authority's attempt to standardise systems, and the lawsuits and criminal fraud trial that followed. 'Centralised systems' has been a dirty phrase ever since.

Despite these setbacks, the NHS's shared services initiative is starting to centralise some back-office functions. University Hospital Birmingham NHS Trust is running a trial of what will become the world's biggest payroll and human resources system. Centralised purchasing and electronic commerce are in the pipeline.But the NHS is primarily about frontline services, and introducing IT here will be much more difficult, for several reasons. One is that the big US vendors find it hard to get their heads round the UK way of providing health. Most countries in continental Europe have health services funded by the state, but not through direct taxation. As a result, there is rarely a ready-to-import solution available.

At the same time, most specialist healthcare IT vendors in the UK are small companies, unable to export their work.

Another big problem is the NHS's federal structure. Hospitals are controlled by primary care trusts, many of which were created only in the latest health service reorganisation. GPs are contracted by the NHS and all have their own ways of working. This makes single procurements of health IT systems very difficult.

On top of this, many healthcare practitioners, particularly senior consultants, tend not to welcome change. Rightly or wrongly, they are inclined to resist the latest big government idea to make them more efficient.

Finally, health records multiply citizens' normal concerns about data privacy several times over. Computing last week reported on a national email system based on PKI which promises to offer the foundations of a secure NHS broadband network. But it will take a long time to convince the sceptics of its virtues and just one error could set back the idea for years.

The expansion of NHS spending announced in the Budget includes a multi-billion-pound national implementation plan for health service IT. This will use a small number of vendors, working with the NHS Information Authority - its central IT department - to computerise everything from individual health records to appointment-booking and automated repeat prescriptions.

Such a programme could revolutionise the NHS. It could eliminate waiting lists. It could win Tony Blair the next election on its own. On the other hand, if the NHS's immune system rejects such an implant, it could be just enough to lose it for him.

Project Watch: NHS DirectAt least the government can point to one success story. Like First Direct, the UK pioneer of telephone banking, NHS Direct was conceived as a way for central management to bypass its unwieldy branch network and get straight to its customers over the telephone and on the internet.

Last January, the public spending watchdog the National Audit Office (NAO) did something almost unheard-of: it offered enthusiastic praise for a technology project, in the form of NHS Direct.

The NAO said the service works safely, with just 29 adverse events in three years, representing less than one phone call out of 220,000. It has also achieved its target of 60 per cent awareness in England and Wales.There are, however, still a few problems.

NHS Direct has a disproportionately white, middle-class, middle-aged user base. And it isn't meeting its targets for answering queries - although if you're on this service's waiting list, you will be seen to in less than an hour, rather than in days or months.

Why has NHS Direct succeeded? It was introduced nationally, bypassing the health service's federal structure. It was a 'green-field site' which didn't involve integration work, so often the stumbling block for any IT implementation. And it was a new service, so its workload could build up gradually.

This consumer-focused frontline is set to become far more important. An email service will be introduced later this year. Next year, the service will become the health service's gateway for all non-English speakers, thanks to the relative ease of introducing translation facilities. And in 2004, the NHS Plan sees the service as the after-hours answering service for GP surgeries. Pilot work in the North-East suggests doctors will receive fewer and more relevant night-time call-outs.

Perhaps NHS Direct's advantage is that it is in tune with public expectations. When the health service was set up, waiting for an appointment was just the way things worked. Half a century later, people are used to being able to simply pick up the phone or look at a web site.

NHS Direct faces enormous challenges as its userbase expands but it has advantages over so many of the other projects featured in this report: clear objectives, realistic timetables and a brief that fits in with today's consumers.