How the national programme came to be the health service's riskiest IT project

NHS IT The potential benefits of the £2.3bn NHS IT plan are huge - but so are the risks. Tony Collins, co-author of a book on the...

NHS IT The potential benefits of the £2.3bn NHS IT plan are huge - but so are the risks. Tony Collins, co-author of a book on the lessons learnt from major projects, reports

At a private meeting at 10 Downing Street in February 2002, officials at the Department of Health made a presentation to prime minister Tony Blair on what would become the UK's biggest civil computer programme.

The proposal was to modernise the health service by spending billions on IT; and it could hardly have had more influential sponsors.

In the room were many great minds who knew a great deal about setting policy but, as one of the senior figures at the meeting said later, not all had a deep knowledge of IT or the NHS.

The attendance list included the chief secretary of the Treasury, the secretary of state for health, and the chief executive of the Office of Government Commerce whose officials oversee major IT projects in government. The e-envoy, who wants to make all government services available online by 2005, and representatives of the Wanless Review Team, who had wanted spending on information and communications technology in the health service to be doubled immediately, were also there.

Blair chaired the event - perhaps the first time a prime minister has conducted a seminar on IT.

By the end of the meeting all were apostles of a centralised approach to IT. One reason was that they wanted to stop hospital staff and GPs wasting hundreds of millions of pounds on computer systems that did not talk to each other.

The seniority of the sponsors ensured there was top-level political and financial commitment to a Whitehall-led IT programme for the NHS. But what if their thinking was wrong? Even if they had the courage, would any of the managers and politicians charged with implementing the centralised IT strategy have the power to reconvene a seminar chaired by Blair to explain that his visionary programme needed reviewing, rethinking, scaling back or even stopped? It seems unlikely.

There was a further problem. The centralised approach, leading to national systems, would require big, ambitious IT projects. And many in the room at Downing Street were aware of a plethora of over-ambitious schemes, led by Whitehall officials, that had ended in disaster. Systems to handle caseworking files for the Immigration Directorate Service had been abandoned; several costly projects had been abandoned at the Department of Work and Pensions; the Lord Chancellor's Department and the Foreign Office too.

Yet those around the table - briefed by major IT suppliers - were convinced that this time it would be different. The then health minister Lord Hunt believed that pace and spread of technological advances, together with the reducing IT costs, had turned technology into a routine tool in the delivery of public services.

Compared to the past, technological solutions were much more readily available to bring about the integration of information and systems.

Hunt also believed that the NHS had learnt from the past not to launch projects that were too big; rather do it by measured stages.

Officials who were at the Downing Street meeting were prepared to concede that the centralised approach to buying IT flew in the face of the government's efforts to devolve power and decisions to trusts and GPs. But the benefits of national systems were too great to ignore.

Blair was told how everyone could have access to their health record. The future was a world in which people booked appointments with their surgeries online. GPs ordered tests electronically and a decision-support system helped the physician choose the right drugs. Hospital staff accessed electronic health records, which were checked and updated and test results and X-rays were transferred electronically across departments.

As important as any of this, researchers looked at records en masse to see what treatments had the best results, and what drugs had the most serious side-effects.

Some of the people who watched the programme evolve saw other less obvious advantages. The national systems could facilitate the introduction of a national identity number so that patients could be authenticated when they accessed their health records; the systems could help track asylum seekers through their medical records, identify benefit and housing fraud, find people in public office with psychiatric histories, or who are paedophiles, or help in the fight against terrorists.

By March 2002 Blair and the chancellor, Gordon Brown, had approved plans for modernising the NHS using IT. They allocated at least £2.3bn for health service IT, on top of the extent £850m a year spend on NHS systems and staff.

A centralist's dream, the Department of Health's National Programme for IT was born.

Inconsistent start

With a Churchillian optimism and doggedness that was to characterise leadership of the national programme for IT in the NHS, Hunt announced details of the plans in June 2002, four months after the seminar at Downing Street.

"It is time we accept that IT is one of those core functions that should be managed and undertaken centrally and get on with it" he told the annual conference of the Association of ICT professionals in Health and Social Care.

At the heart of the national programme for IT were four national projects: a new broadband-based infrastructure, and three "critical" national services of e-health records, e-prescribing and booking of hospital and GP appointments online.

If nothing else were delivered and available on a national basis these four projects would be, he said. But nobody noticed an inconsistency in his speech. The government had learnt from past failures to baulk at projects that were too big or ambitious, he said; but in another part of the speech he went on to emphasise the ambitious nature of the national programme.

"I hope you will agree that this programme and implementation plan is ambitious, because it is meant to be," said Hunt.

Risk scores

As a test of whether the design of a nascent project was overly ambitious, civil servants devised a score card of risks.

Points from one to four or one to six were awarded according to the risks in particular categories, six being maximum possible risk. For example, if a project cost less than £10m, affected fewer than 1,000 people, used proven technology, had no significant impact on the organisation, and needed no interfaces with existing systems it would gain only five points, the lowest possible risk.

But in these particular categories the national programme for IT scored close to the maximum risk: 20 out of a possible 23 points, according to an internal policy document of the Department of Health which is marked "Restricted - policy".

When the scores in these categories and other categories in the test were added up, the national programme scored a total of 53 out a maximum of 72.

The high risk score arose because the programme was said in the document to cost more than £100m, affect more than 10,000 people in the NHS, the technology or scale of its planned use was unproven, -and it was dependent on significant new business processes which required new skills.

The document, which is an unpublished appendix to a public report Delivering 21st Century IT Support for the NHS also said that the NHS and some suppliers were inexperienced in the use of the required technology, and there would be significant challenges in converting existing data to run on new systems, and in seeking to integrate legacy applications with new ones.

But the score sheet omitted one major risk that has been a significant factor in the failure of many IT projects: a short time scale. Politicians insisted that the national programme for IT was implemented quickly. So today, although many observers say the risks of implementing the national systems have not abated, there is little opportunity for a rethink.

The timetable for contracting out delivery of systems under the national programme to external companies seems non-negotiable. The first contracts must be signed by 31 October, the rest by the end of the year. If contracts are not signed by Christmas, there will be no Christmas, said Richard Granger, director-general of NHSIT, who joined the Department of Health from the private sector in October 2002 to put into practice the principles agreed at the Downing Street seminar.

The Department of Health's unyielding approach to the timetable was set out in its publication Key Elements of the Procurement Approach to the national programme in January. It listed speed as its first key requirement. "Rapidity: this means the commitment of all parties to do things quickly," said the document.

The political pressure to complete the implementation of national systems quickly existed long before Granger was appointed to run the programme.

One of those who attended the February meeting at Downing Street was Professor Sir John Pattison, a senior executive at the Department of Health. Afterwards he discussed the national programme with colleagues who welcomed the proposals.

"There was only one question which I thought was rather tricky and that was 'How long will this take?'" said Pattison. "I swallowed hard because I knew I had to get the answer right for the purposes of the audience in which I was standing and I said three years. The answer was: 'That is too long, how about two years?' but in the end we got two years and nine months, starting from April 2003."

But an artificially short time scale has been a factor in some of the UK's most damaging IT and other disasters. A public inquiry in 1993 into the collapse of a computer-aided despatch system for London Ambulance Service determined, for example, that a rushed procurement process contributed to the system's collapse.

The ambulance staff union claimed that the system's failure cost 11 lives because ambulances were delayed in reaching patients because of problems with the new system.

The official report into the collapse of systems at London Ambulance said of the procurement: "The size of the programme and the speed and depth of change were simply too aggressive for the circumstances." The report also said that management had "ignored or chose not to accept advice provided to it from many sources outside of the service on the tightness of the timetable or the high risk of the comprehensive systems requirement".

In recent months, some prospective suppliers to the national programme for IT in the NHS, and others, have expressed astonishment at the speed with which the Department of Health is moving toward signing contracts with private companies to deliver new systems under the programme.

According to a confidential document published internally by the Department of Health, and seen by Computer Weekly, some bidders have expressed concern about the speed of the procurement.

Angus Goudie, IT head for the Sunderland Teaching Primary Care Trust, is among the doctors who claim the national programme is being rushed. He says, for example, that some clinicians are unable to attend demonstrations of new systems because they are given too short notice. "We want to see the national programme work well but we do feel that the advisory stage with clinicians is being rushed."

Suppliers report regularly that they are having to respond in detail to specifications that change frequently; and clinicians say that they are setting aside time they could spend with patients to see demonstrations of systems only to find that some of these events are cancelled at short notice.

But a spokesman for the national programme denied that the negotiations for choosing suppliers were being rushed. "The national programme is running a procurement process to a challenging, but achievable, timetable."

Forward, unflinching, unswerving, indomitable, till the whole task is done, said Churchill in May 1945 as he prepared the nation for the continuing war against Japan. The Department of Health sees itself as if in a war against what everyone agrees is the generally poor state of NHS systems. Risks must be taken and casualties expected if the enemy of inefficiency is to be defeated and the benefits realised.

But this approach was adopted by managers of the doomed project to automate the despatch of ambulances to emergencies. The report of the public inquiry into the failure of the ambulance systems said that management "created an atmosphere of mistrust" with an over-aggressive style born in part out of the desperation to put right decades of poor performance.

Nobody could quibble with the philanthropic aims of the NHS's national programme. If it works well, patients can only benefit. Lives may even be saved if doctors and nurses have ready access to comprehensive and regularly updated medical records. But there are high risks of huge losses, disruption to the work of clinicians and, worst of all, an enormous opportunity lost.

So are the risks worth the possible benefits?

Clearly Blair thinks so, as do representatives of major suppliers and officials at the Department of Health who are also sincere in their commitment to improving the lot of patients.

But even if those at that all-important February meeting at Downing Street had any reason to believe that the national programme for IT was too high risk to continue without stopping for a major review, would they - or could they - give the bad news to the chief executive of the board, Tony Blair?

Most senior politicians, one suspects, like to hear only good news. This could create an unbridgeable divide between those who set the policy and those who have to implement it.

One group of nurses has likened the national programme for IT to an unstoppable juggernaut. Critics could stand on the sidelines waving their arms and warning the drivers to slow down; but they are unlikely to be heard by the programme's apostles who appear to regard all their detractors as Luddites.

Meanwhile the innovatively designed national programme for IT tears ahead, proof that the government is bold enough to invest huge sums of public money into potentially one of the UK's biggest and riskiest computer projects, for good or ill.

Richard Granger, director-general of NHS IT, comments on the NHS' national programme   

We all need to be engaged in an IT programme of major importance to the NHS and this is our one and only chance in this generation to support change and service improvement through the better use of information and IT. 

There is an assertion by some commentators that the national programme is broken before it begins. This is disingenuous, inaccurate and unfounded. 

The procurement timetable is challenging but necessarily so. The NHS is expecting the private sector to commit substantial high-quality resources to the bidding process. A vibrant procurement brings with it the benefits of maintaining a management focus, better engagement with prospective suppliers and the framework for ensuring patients and end-users can experience tangible change for the better more quickly. 

Our approach to procurement could be described as "intrusive supplier management". This entails the national programme for IT providing the vision and defining the objectives. Then drafting and refining the contract to ensure the contract can be used as a sound basis for managing the commercial activity and the framework for delivery. 

We are breaking down the national programme into manageable chunks based on function, geography and phasing. It will be an incremental approach that over time builds on each phase of delivery and functionality.  Finally, several other key strands of work are well under way, including improving clinical and patient involvement and developing the right approach on confidentiality.

How will the NHS measure the success of the national programme for IT?   

Healthcare professionals including Frank Burns, chief executive of Wirral, one of the most technologically advanced trusts in the UK, said success will be determined by whether new national systems are used to good effect by doctors and nurses. 

"All my experience tells me, and I think that is the view of many people, that local clinical ownership is the most critical success factor in ensuring successful implementation of clinical information systems," Burns said.  

"But I am honestly not sure that getting high levels of local clinical ownership is remotely possible with procurement decisions that are covering half the country, and I think that we have to recognise that problem, and I think that we all have a huge selling job to do."  Some doctors and nurses are concerned that the Department of Health will judge the programme a success even if systems are not used effectively.  

This is because the Department of Health may choose to measure success according to the number of doctors, nurses and administrative who have access to - rather than use - new national systems. If this is the success criteria, the national programme cannot fail.  

IT suppliers working at computer sites that are external to the NHS need only provide new equipment and access points to these new systems within hospitals, GP surgeries and other health centres. Then, theoretically, hundreds of thousands of doctors, nurses and administrative staff will have access to the new, national systems, but may use them primarily only for mandatory functions that have little to do with improving the care to patients.   

The national programme will be a genuine success, however, if doctors and nurses across the NHS use the new systems and use them because they are unequivocally beneficial to them and the care of their patients. But this will be hard to achieve.  Burns said getting clinicians to feel part of the programme is potentially the biggest risk factor in the national programme.    

He added, "I think the issue needs a lot more attention than I am aware that it is currently getting."

How the national programme plans to mitigate the key risks     

Risk: A lack of independent assessment of the programme 

Solution: The national programme and implementation plan for IT will be subject to the Gateway process with the Office of Government Commerce. The Gateway process is designed to improve the management of major IT activity involving procurement or partnering. It applies best practice to support the successful delivery of programmes and provides an early review and continuing independent assurance.   

Risk: Lack of co-operation and buy-in by clinicians and other stakeholders to objectives of investment 

Solution: Ensure full involvement of interested parties. Also the Department of Health will provide clear leadership   

Risk: Services are not affordable  Solution: Scope the programme accordingly    

Risk: Individual organisations within the NHS act unilaterally 

Solution: Clear leadership and set targets through performance management   

Risk: Shortage of required local skills 

Solution: Minimise by transferring risk to private sector partners, identifying high-calibre staff within the NHS and ring-fence required staff and skills   

Risk: Problems migrating from current systems 

Solution: Adopt a clear communications strategy to facilitate local planning, and ensure adequate transition planning   

Risk: Costs escalate 

Solution: Ensure that investment has a ceiling, there is strong budgetary control, clearly defined review points, regular monitoring of costs and exception reporting   

Risk: Costs cannot be predicted 

Solution: Develop a cost model as part of the negotiation of contracts with prospective suppliers.

Canadian case shows that clinicians need to be included 

Healthcare specialists who helped to implement a flagship system to handle patient records in Toronto, Canada, have warned that the UK's national programme for IT could fail unless clinicians feel involved in the projects from the outset. 

Matthew Morgan, a physician, former head of clinical informatics at University Health Network and now head of informatics at Per-Se, a US-based specialist in electronic patient records, said he has studied the NHS' national programme for IT, and he has some concerns. 

"Getting every clinician's perspective prior to implementation is not possible but ensuring local clinician leadership is possible and is essential to success. Regardless of how extensive the consultation process has been at a national level, it has to be implemented locally. Without this it will fail. 

"To get local clinician champions you have to be prepared to understand and solve local issues that clinicians have with the Integrated Care Records [patient records] Service.  

"The question is, do you want to tackle those issues in a constructive positive environment prior to the implementation or do you want to take a chance and risk dealing with clinician dissatisfaction and rebellion at the post implementation stage?" Morgan added. 

A spokesman for the national programme said, "We value the input of clinicians and other NHS staffÉ We are pleased to be already receiving extensive input from more than 100 clinicians, which is helping us to drive the programme forward." 

Other supporters of the national programme say, however, that the challenge will be gaining the buy-in not of hundreds of clinicians but the tens of thousands who work in hospitals and GP surgeries.   

Len Slawinski, managing director of Per-Se in the UK, said the NHS may be too big to treat as a single entity for the purposes of integrated care records. He said a national data spine should be installed but he wants trusts to have a choice in selecting their local systems.  

"There is no point in introducing the clinicians late into the selection process after the local service provider has been appointed and after that provider has established its product supplier list. By then and clinicians will be faced with a fait accompli."

Read more on IT for government and public sector