Information Technology (NHS) Debate
Full Debate: Read Full DebateRichard Bacon
Main Page: Richard Bacon (Conservative - South Norfolk)Department Debates - View all Richard Bacon's debates with the Department of Health and Social Care
(13 years, 2 months ago)
Westminster HallWestminster Hall is an alternative Chamber for MPs to hold debates, named after the adjoining Westminster Hall.
Each debate is chaired by an MP from the Panel of Chairs, rather than the Speaker or Deputy Speaker. A Government Minister will give the final speech, and no votes may be called on the debate topic.
This information is provided by Parallel Parliament and does not comprise part of the offical record
It is a pleasure to serve under your chairmanship, Dr McCrea. I am reminded of Fidel Castro’s old maxim that any speech of less than three or four hours cannot be any good, and when I reviewed the material available for this debate, I felt that it might be difficult to put everything I want to say into a shorter time. I appreciate, however, that this debate lasts only 90 minutes and that the Front-Bench speakers will be called fairly soon. I am not planning to take many interventions during my remarks, so that I can get through everything I wish to say as quickly as possible.
The importance of information technology to good health care cannot be overstated. As leading health informatics expert, Dr Anthony Nowlan, put it:
“Redesigning the ways care is organised and conducted and supporting those new ways with information science is more important to people’s health overall than any new drug we could develop in the next decade.”
He also stated that
“the engagement of clinicians and managers is not just about telling them what is going to happen.”
Sadly, those words accurately summed up a significant part of the problem that we faced.
The national programme for IT in the health service is the largest civilian computer project in the world. It was spawned in late 2001 and early 2002, after the then Prime Minister, Tony Blair, met Bill Gates and was bowled over by a vision of what IT could do to transform the economy and health service. The idea was for information to be captured once and used many times, transforming working processes and speeding up communications. A far-reaching vision set out a programme that would supposedly lead to a transformation in people’s experiences of health care. Hospital admissions and appointments would be booked online—the choose and book system—pharmacists would no longer struggle with the indecipherable handwriting of GPs; and drug prescriptions would be handled electronically. There was to be a new broadband network for the NHS, a new e-mail system, better IT support for GPs and digital X-rays. Most important of all, medical records would be computerised, thus transforming the speed and accuracy of patient treatment through what became known as the NHS care records service.
The NHS care records service comprised two elements: first, a detailed care record that contained full details of a patient’s medical history and treatment. That was to be accessible to a patient’s GP and to local community and hospital care settings, so that if treatment were required, all the information would be available. Secondly, there would be a so-called summary care record that contained medical information about things such as allergies and would be more widely available.
It became clear that Tony Blair was in no mood to wait when he asked Sir John Pattison, who attended the Downing street seminar in February 2002 where these matters were discussed, how long the IT programme would take. Sir John Pattison later stated:
“I swallowed hard because I knew I had to get the answer right… and I said three years.”
Tony Blair replied, “How about two years?” and they settled on two years and nine months from April 2003—in other words, until December 2005. Given the extent of the proposals, that was a ludicrous timetable. Nevertheless, the decision had been made, and everything had to be done at breakneck speed.
Sir John Pattison and his team set to work and produced a blueprint entitled “Delivering 21st century IT support for the NHS: national strategic programme”, which was published in June 2002. The aim was to connect the delivery of the NHS plan with the capabilities of modern information technology. There was, however, an odd discrepancy at the outset. At the back of the original document were four appendices, one of which contained the project profile model and stated that the project’s estimated whole-life costs were £5 billion. It provided a total risk score of 53 out of a maximum of 72. In other words, the project was very high risk. When the document was published, however, that project profile model had been removed and there were only three appendices—the likely costs of the project and the true risks were concealed right from the start. After the publication of the document, the Department of Health established a unit that later became the Connecting for Health agency. In September 2002, Richard Granger was appointed as director general of the NHS IT programme on a salary of about £250,000. His job was to turn the national strategic programme—which soon became the national programme for IT in the health service—into reality.
Richard Granger was a former Deloitte consultant who had successfully overseen the introduction of the London congestion charge. Speaking at a conference in Harrogate some months after his appointment, he announced that the cost of the IT programme would be £2.3 billion. That figure contrasted with the unedited version of Sir John Pattison’s “Delivering 21st century IT support for the NHS”, which a few months earlier had come up with the larger estimate of £5 billion.
Mr Granger commissioned a study by McKinsey into the health care IT market in the UK, which was then dominated by medium-sized firms that sold systems to hospitals and GP surgeries. The study concluded that no single player was capable of becoming a prime contractor in a multi-billion pound programme, and Mr Granger soon announced that the procurement process for the programme would be structured to attract global IT players. He had little respect for the skills of most public sector buyers of computing systems—perhaps with good reason if one looks at the track record—and knew that IT contractors routinely run rings around their customers in government.
Mr Granger made it clear that things would be different on his watch. Contractors would not get paid until they delivered, and those not up to the mark would be replaced. He even compared contractors to huskies pulling a sled on a polar expedition:
“When one of the dogs goes lame, and begins to slow the others down, they are shot. They are then chopped up and fed to the other dogs. The survivors work harder, not only because they’ve had a meal, but also because they have seen what will happen should they themselves go lame.”
Mr Granger started as he meant to go on, and potential contractors were left in no doubt that the procurement process was to happen quickly. In May 2003, potential bidders were given a 500-page document called a draft output-based specification, and told to respond within five weeks.
One of the classic failures in many IT projects is the failure to consult adequately with those who will use the systems once they are delivered. The national programme followed that pattern in many respects, but in this case that did not happen by accident. Mr Granger had no patience with what he saw as special pleading by medical staff, whom he believed were unwilling to accept the ruthless standardisation that was necessary to deliver the advantages offered by the IT system. He effectively believed that he knew what the clinicians needed better than they did themselves.
Some clinicians were keen to ensure that they had proper input into what was happening. Sir John Pattison asked Dr Anthony Nowlan, the health informatics expert who at the time was the executive director of the NHS Information Authority, to secure the involvement of health professionals in the programme. The aim was to obtain a professionally agreed consensus about what was the most valuable information to store, and what was achievable in practice.
After several months the group had hammered out a consensus, but although that work was fed in when the contracts began to be specified, it formed only a relatively small part of the overall specification. The large majority of the so-called output-based specifications, and the crucial major hospital systems at the heart of the programme, were developed without involvement and scrutiny by the leadership of the health profession. That happened despite the fact that involvement by users is essential if one wants software that works and that people will use.
The great speed at which contracting was completed meant that all complex issues had to be faced after the contracts had been let. Anthony Nowlan began to realise that his efforts were not welcome, and he told the Public Accounts Committee that
“it became increasingly clear to me that efforts to communicate with health professionals and bring them more into the leadership of the programme were effectively obstructed.”
Worse still, Nowlan was subsequently asked to provide a list of the names of hundreds of people who had been involved in specification work, so as to provide evidence to reviewers that the work was valid. In fact, all that had happened was that an e-mail had been sent out. Quite understandably, Dr Nowlan thought that saying that people had been consulted because they had been sent an e-mail was not consultation in any proper sense, any more than compiling a list of people who had been sent an e-mail was proper validation. He regarded the claims as a sham, and refused to co-operate.
It turned out that serious clinical input into the programme was not really wanted. As Professor Peter Hutton later told the PAC,
“it was like being in a juggernaut lorry going up the M1 and it did not really matter where you went as long as you arrived somewhere on time. Then, when you had arrived somewhere, you would go out and buy a product, but you were not quite sure what you wanted to buy. To be honest, I do not think the people selling it knew what we needed.”
The result was a set of contracts that were signed before the Government had understood what they wanted to buy and the suppliers had understood what they were expected to supply.
When the then Health Secretary John Reid—now Lord Reid—announced the contract winners in December 2003, the value of the contracts had already shot up to £6.2 billion from the original £2.3 billion mentioned by Mr Granger in Harrogate. The time scale had tripled in length, and instead of the two years and nine months from April 2003 originally promised—to which Sir John Pattison had been obliged to commit at the Downing street seminar—the contracts were now to run for 10 years. Later, one of the most senior officials in the national programme, Gordon Hextall, even claimed that it was always envisaged that the programme would run for 10 years.
Four winning bidders were appointed: Accenture; Computer Sciences Corporation, or CSC; Fujitsu and BT. They were known as local service providers, or LSPs. BT and Fujitsu picked a US software firm, IDX, to work with, while Accenture and CSC both picked a British software company called iSoft. iSoft was a stock market darling that had been spun out of the consulting firm KPMG in the late 1990s. The company’s flagship was a software system called Lorenzo, which was portrayed enthusiastically in iSoft’s 2005 annual report and accounts. The chairman, Patrick Cryne, told shareholders that Lorenzo had made “impressive progress”, while chief executive Tim Whiston stated that Lorenzo would be “available from early 2004” and that it had
“achieved significant acclaim from healthcare providers”.
With such promising statements from the company’s directors, the stock market was delighted, and it was no surprise that iSoft’s share price rose sharply. Mr Cryne, Mr Whiston and their fellow directors then sold large tranches of their personal shareholdings in iSoft, making around £90 million in cash. In 2004, Patrick Cryne bought Barnsley football club.
There was a slight problem. The flagship product, Lorenzo, which was described in such encouraging terms by the directors, was not finished. That caused a big headache for Accenture, the biggest LSP, with two contracts worth around £1 billion each. It was in partnership with iSoft and was trying to implement software that was basically not implementable. CSC faced a similar problem in the north-west. Under the Granger rules of engagement, no one was supposed to get paid until something was delivered. As iSoft had not produced a working version of Lorenzo, the brutal reality was that neither Accenture nor CSC had any software to deploy.
There were still big concerns about the programme’s indifference to securing clinical buy-in from users—clinicians in hospitals—even though numerous studies had pointed to such buy-in as the key ingredient for success in any IT project. Professor Peter Hutton wrote to the then chief executive of the NHS, Sir Nigel Crisp, to express his continuing disquiet:
“I remain concerned that the current arrangements within the programme are unsafe from a variety of angles and, in particular, that the constraints of the contracting process, with its absence of clinical input in the last stages, may have resulted in the purchase of a product that will potentially not fulfil our goals.”
Soon after pointing out politely that the emperor had no clothes, Professor Hutton was asked to consider his position, and he tended his resignation. The IT people were simply not interested in what the doctors were telling them. To give it belated credit, however, the Department of Health began to realise that securing the support and buy-in of clinicians who would have to use the systems might be a good idea.
In March 2004, the deputy chief medical officer, Professor Aidan Halligan, was appointed alongside Richard Granger as joint director general of NHS IT, and joint senior responsible owner of the programme, with specific responsibility for benefits realisation. That was welcomed by clinicians. One delegate at the Healthcare Computing conference in Harrogate said that Halligan’s appointment was “really, really good” because
“he has the trust of clinicians and can stand up to Granger”,
although a general practitioner delegate at the same conference said that it “spoke volumes” that nobody like him had been in the post earlier. Halligan acknowledged that not enough had been done to win the support of clinicians, whose buy-in, he said, was critical to the success of the project. Listening to clinicians was now the flavour of the month. However, there was one insuperable difficulty—the contracts had already been signed. As Professor Hutton later explained to the PAC,
“it became clear from discussions with suppliers in early 2004 that what they had been contracted for would not deliver the NHS Care Record”.
Accenture and CSC struggled on with the unusable Lorenzo. Eventually they commissioned a study that produced a confidential report in February 2006, which confirmed their worst fears. The report stated that the Lorenzo had
“no mapping of features to release, nor detailed plans. In other words, there is no well-defined scope and therefore no believable plan for releases.”
That was over five years ago.
In March 2006, Accenture announced to its shareholders that it would use $450 million to cover expected losses on the programme. It made repeated offers to the programme that it would meet its contractual obligations by using other software. However, that might have bankrupted iSoft, and Richard Granger was having none of it. He responded with a threat when Accenture talked about walking away. Referring to tough penalty clauses contained in the contracts, he said that
“if they would like to walk away, it’s starting at 50% of the total contract value”.
Accenture had two of the £1 billion-a-piece prime contracts, so it appeared to be facing a cool £1 billion in penalty payments to the Government if it abandoned the programme. Strangely, it did not work out that way. Accenture engaged in swift negotiations with the health service and in September 2006, after making a penalty payment of just £63 million, it duly exited the programme. Mr Granger’s threat, that if Accenture left the programme it would face gigantic penalty payments, proved to be of little account. There were rumours that if Mr Granger had demanded any more money, he would have faced serious and embarrassing counter-claims from Accenture for failures by the national programme to stick to its own contractual obligations.
CSC, with its own £1-billion contract for the north-west and west midlands regions, was in no better a position than Accenture to implement the unfinished Lorenzo software. It was also struggling to mop up after having caused the largest computer crash in NHS history, when its Maidstone data centre was hit by a power failure, followed by restarting problems. The back-up systems did not work, and data held in the centre could not be accessed. That meant that, for four days, 80 NHS trusts could not use their patient administration systems and had to operate as best they could with paper systems.
Another worry for CSC was its shareholders. Accenture had set aside hundreds of millions of pounds against expected losses and told the stock market accordingly, but CSC had done no such thing. In addition to its problems with losses in the UK, the company had troubles back home in the United States, where it faced allegations of corruption. The US Department of Justice had alleged that CSC was part of an alliance, which included virtually all the major sellers of hardware and software in the United States, that had swapped unlawful kick-backs in Government agency technology contracts. CSC finally agreed to a $1.37 million payment to resolve those allegations. That was reported on the news blog of Cnet.com on 13 May 2008, under the heading:
“CSC settles with feds over kickback allegations”.
In such circumstances, having extra contracts from the NHS might look reassuring to the US stock market. Despite the fact that there was no implementable software—Lorenzo still was not finished—CSC quickly took on both Accenture contracts, tripling its involvement in the programme. However, there were continuing problems at iSoft, which was supposedly writing the Lorenzo software. One of the problems related to the publication of iSoft’s financial results, which had been repeatedly delayed, up to the point where one of iSoft’s own advisers, Morgan Stanley, a brokerage, declined to publish a profit forecast, stating:
“We don’t feel we have enough visibility to offer a recommendation”.
With friends like that in the stock market, who needs enemies? Finally, iSoft was forced to declare a loss of £344 million, which wiped out all the company’s past profits. The Financial Services Authority launched an investigation.
Now, three fifths of the programme was dependent on one troubled local service provider, CSC, which was using a software supplier, iSoft, that was itself under investigation by the FSA. One regional contractor, Accenture, had been replaced by another, CSC, which had less experience. The central problem remained: the software that they had been trying to deploy, iSoft’s Lorenzo system, was still not finished.
In those circumstances, iSoft started to deploy software products that predated the programme, which Connecting for Health duly paid for. Those older products did not meet the specifications for the national programme. It is important to remember that fact, because that is what many acute hospitals have now been given—old and outdated software that was deemed inadequate nine years ago to meet the programme’s specifications.
Meanwhile, the other two providers, BT and Fujitsu, were having their own problems. They were trying to implement American software, which is not such an easy thing to do in a British hospital, because American hospitals rely on billing for each and every activity and do not, conversely, expect to have to handle waiting lists. An American software system cannot be just uploaded to an acute hospital main frame and be switched on—it is not that simple.
In June 2005, IDX was dropped by Fujitsu with Richard Granger’s consent and replaced by another American firm, Cerner, which had a software package for large acute hospitals called Millennium. BT, some 18 months after winning its LSP contract, was still struggling with IDX. By July 2005, BT was facing serious threats from Richard Granger that it could be axed if it did not start to perform. In an interview with Computing magazine, Mr Granger said:
“BT had better get me some substantial IDX functionality by the end of summer or some predictable events will occur.”
However, it was not that simple. As the leading health care IT website, e-Health Insider, pointed out, replacing BT as the local service provider
“would represent a major failure for the programme, and raise questions over the whole IT-enabled NHS modernisation”
and lead to even more delays. The website added:
“Such a move would also potentially raise serious questions about whether the adversarial management style of Connecting for Health is the most likely to deliver new systems that provide clinical benefits to patients in a timely and cost-effective fashion.”
BT was allowed to continue as the local service provider and eventually, with Granger’s consent, it was allowed to follow Fujitsu’s lead and replace IDX with Cerner Millennium.
At a London conference in July 2005, Mr Granger gave a stern warning to suppliers who were lagging behind on delivery:
“We will get very soon to a point where they will either come good with what they’ve got, or they will get a bullet in the head.”
Mr Granger was also showing signs of defensiveness about the programme, stating:
“It might be a policy disaster, but it isn’t an IT disaster. The system was delivered to spec”,
and he gave the example of the electronic staff record. He added:
“If some of my colleagues do not think sufficiently through as to what was wanted then it’s a specification error.”
Such statements by Mr Granger led to howls of rage from some industry observers, including one who, after Granger’s speech, posted a comment on the e-health Insider website, saying:
“Now and then I check myself from hatred of what Richard Granger stands for and has done to NHS IT, and then the sheer arrogance and ignorance of his public statements brings me back. He set the ridiculously short timescales for decision-making, procured before there was a clear idea of the scope, handed all the ‘choice’ from NHS clinicians to private contractors. CfH”—
Connecting for Health—
“hasn’t solved the funding crisis for computerising the NHS, rather landed us with a massively expensive way to do what some of us were achieving already”.
Meanwhile, the National Audit Office had embarked on a study of the national programme, which was due to be published in summer 2005, but there were considerable delays. As Members may know, NAO reports involve a clearance process, during which a report’s factual content is cleared with the Government before publication, and that has benefits for both sides. However, something different happened with the national programme report. It was as if Connecting for Health wanted to use the clearance process to expunge the slightest criticism of its activities. It undertook a war of attrition with the auditors, in a process that the NAO later described as fighting
“street by street, block by block”.
The final report was delayed again and again, and it finally appeared in June 2006. It was much weaker than seasoned health IT observers had expected. The Minister of State, Department for Communities and Local Government, my right hon. Friend the Member for Tunbridge Wells (Greg Clark), who was then a member of the PAC, described it as “easily the most gushing” he had read, while a BBC correspondent described it as a “whitewash”. Most of the key criticisms were eventually excised, as Granger and his team ground down their opponents. It later emerged through freedom of information requests that earlier drafts had been much tougher.
Tom Brooks, a management consultant with years of worldwide experience in health care, wrote a devastating submission to the PAC, in which he questioned the whole rationale for central procurement in the programme. He said that
“the poor quality of the negotiation of the NPfIT contracts by Mr Granger”
was a subject of criticism. He described the view that central procurement would produce systems that met local requirements as “a fundamental error”. He told the Committee:
“MPs are mis-informed if they view the central infrastructure as making reasonable progress”.
Dr Anthony Nowlan, whom I mentioned earlier, described the programme as “back to front”, given that the contract stating what would be produced had already been let. He pointed out the sheer absurdity of a consensus document produced by the programme stating:
“Now that the architecture for England has been commissioned, designed and is being built, there is a need for clarity concerning how it will be used”.
A group of health IT experts sent the PAC a detailed paper offering a devastating critique of the entire programme. The group provided evidence that it was likely to deliver neither the most important areas of clinical functionality nor the benefits required to justify the business case. The group simply stated:
“The conclusion here is that the NHS would most likely have been better off without the National Programme in terms of what is likely to be delivered and when. The National Programme has not advanced the NHS IT implementation trajectory at all; in fact, it has set it back from where it was going”.
In view of the frequent misunderstandings about the national programme among so many journalists, broadcasters, politicians and commentators, it is worth quoting the expert group’s document at some length. It starts by saying:
“It is useful to begin with the question: What is the central point of NPfIT—its chief raison d’etre? Is it a shared medical record (otherwise known as the ‘Central Spine’ or ‘Central Summary Care Record Service’) across England?
The answer to this important question is simply: no…the central point of NPfIT is to provide the local Care Record Service...Compared with the local CRS, the Central Spine is a much lower priority because it is totally speculative and even if delivered is likely to result in very little clinical benefit…This is a subtle but critical point. The Local CRS systems…are a proven technology…These local CRS systems have always been costly investments (several million pounds per hospital over several years) but have been proven in the NHS and elsewhere to deliver real clinical benefits…This picture is entirely different for the so-called Central Spine record, or Central Shared Summary record, which NPfIT (and the government ministers) would like the public to believe is the central point of NPfIT. It is not. The Central Spine record is just a concept…The problem is that clinicians have told us medicine does not work like this. Clinicians do not just use a summary record to deliver care. They build and depend upon detailed and specific medical data that are relevant for each patient.
They do not rely on some other clinicians’ definition of what will be most relevant to put in a summary record. What is relevant clinically will inevitably vary from patient to patient.
The concept of a summary Central Spine record has no scientific basis and no significant clinical support to back it up—just an overly simplistic and naïve storyline about a Birmingham patient falling ill in Blackpool. In fact, no one has ever provided any figures on how often this situation is likely to arise to show whether or not the investment in the Central Spine record is worthwhile.
The point here is that the Local Care Record Service”—
I emphasise the word “local”—
“is the essential building block for clinically useful health IT to support clinical care in progressive, modern and proven ways. Yes, it is difficult to implement and can take 2-3 years to roll-out across the whole hospital (or organisation), and yet it is always worthwhile…These Local Care Record Service systems are the building blocks and are the point of NPfIT, and what NHS Trust Chief Executives want, need and expect. They are not waiting for a Central Spine record to run their hospitals.
However, the Local Care Record Service systems (or the Local Service Providers’ newest versions of them) are not likely to be fully deployed now (only the rudimentary patient administration elements of them will be) because NPfIT is putting in old ‘legacy’ products in place of new modern Local Care Record Service products in its panic to show deployment and because the systems have been so late in being delivered by the LSPs”—
the local service providers. The document continues:
“The key point of the National Programme for IT is to provide both depth of clinical systems functionality and breadth of integration in terms of delivering the contracted Local CRS functions across organisations and care-settings (acute, primary, mental health, social services).
This is the true vision of health IT promised by the National Programme which is embodied in the Local Service Provider contracts and it is what their price reflects.”
The trouble is, with all the delays, the LSP schedules are being down-scoped behind the NHS’s back and without any accountability to the local NHS Trust chief executives to whom the original vision was promised.”
In September 2006, with the hon. Member for Southport (John Pugh), I published a paper called “Information technology in the NHS: What Next?” In it, we identified four fallacies and offered a way forward. The fallacies were that
“Patient data needs to be accessible all over the country…Local trusts can’t procure systems properly so the centre has to do this for them…Large areas of the NHS need to work on a single massive system”
and that the
“National Programme saves money.”
The suggested way forward was to allow hospital chief executives to buy the systems they actually wanted, subject only to common standards, and to fund such purchases partially from the centre, while making local chief executives contractually responsible for delivery.
Shortly after we published that paper, the NHS chief executive, David Nicholson, introduced the NPfIT local ownership plan, but it did not follow our suggestion of giving local chief executives autonomy in what they bought. Under the NLOP, hospital chief executives would still be required to buy the software that the local service provider was contracted to provide—the difficult-to-install American system, Cerner Millennium, or the non-existent Lorenzo.
Furthermore, instead of there being one senior responsible owner for the programme, which is a central tenet of good project management practice, there would be many dozens of senior responsible owners dispersed among the different primary care trusts, strategic health authorities and hospitals across the country. Those bodies were given responsibility for implementing and delivering software that was not available or which did work properly, without a free choice to buy something else that did work. The NLOP looked more like an attempt to decentralise impending blame than a serious attempt at reform. That is why Tony Collins, one of the country’s leading computer journalists, playfully said that NLOP actually stands for “No Longer Our Problem”.
In February 2007, Andrew Rollerson, a senior Fujitsu manager who had assembled and then led the winning Fujitsu team in the original bid process, mentioned more or less en passant at an IT conference that his view of the national programme was that
“it isn’t working and it isn’t going to work”.
To many informed observers, it was just a statement of the obvious. The PAC called him to give evidence, and when asked if he felt that he had been the
“one who let the finger out of the dam”
and allowed
“a whole collective sigh of relief”
to go round the health IT sector, he replied
“I think that is absolutely spot on.”
Fujitsu then wrote to the Committee stating that Rollerson was not a senior executive of the company and had not been involved for a long time, but neglecting to mention that he had led the winning bid team.
By 2007, another accounting probe had been launched into iSoft by an accounting standards body, and in the following month, April, the PAC published its report, which concluded that
“at the present rate of progress, it is unlikely that significant clinical benefits will be delivered by the end of the contract period”.
By June 2007, Richard Granger had announced that he would quit at some point and shortly afterwards stated that he was “ashamed” of some of the systems put in by Connecting for Health suppliers, singling out Cerner for criticism. David Nicholson, the NHS boss, appointed several new senior executives to join Granger at the top table, while continuing to reject calls for a full review. Tony Collins wrote in Computer Weekly that the future of the national programme for IT in the NHS was “hazy” and that it was becoming
“difficult to delineate success from failure”.
Derek Wanless, whose major review for Tony Blair into the future of the health service had first identified investment in IT as an area for improvement, publicly questioned whether the NHS IT programme should continue without a full audit. He said that
“there is as yet no convincing evidence that the benefits will outweigh the costs of this substantial investment”.
In October 2007, the Department of Health rejected rumours that Matthew Swindells had been appointed interim chief executive of Connecting for Health, but in an industry survey he was named the 12th most influential person in the NHS—10 positions above Richard Granger. It appeared that Richard Granger’s influence was on the wane and that he was being eased out. Tony Collins mused on his blog that the programme might be even worse without Richard Granger—
“the thought of this juggernaut being without a driver is even more scary that when it had a driver but no controls”.
Mr Granger’s last day as an employee of the NHS was 31 January 2008, though, curiously, it was a week, on 6 February, before the interim director of NPfIT and systems delivery, Gordon Hextall, sent a letter to Connecting for Health staff to tell them that Granger had gone and that two appointments would replace him: a top-level chief information officer and a director of IT programmes and systems delivery. Meanwhile, the interim chief information officer would be none other than Matthew Swindells, whose involvement the Department had denied earlier.
In February 2008, the Commons Health Committee published a report on the electronic patient record, which stated that it was “dismayed” by the lack of clarity about what information would be included in the summary care record and for what the record would be used. It also said that there was “a stark contrast” between the “specific and detailed” vision set out for the integrated care records service in 2003 and the “vague and shifting” vision set out in 2007. The Committee concluded that there was now a
“perplexing lack of clarity about exactly what NPfIT will now deliver.”
In May 2008, the NAO published a progress update, which was much more robust than its earlier report. It concluded that the programme has
“largely failed to deliver on its central objective of detailed care record systems for acute hospital trusts”.
Not a lot was happening at that point because there was no software to deploy, so many people were employed but they were not necessarily doing very much. In October 2008, Nick Timmins of the Financial Times wrote about the national programme in a front-page story:
“Progress has virtually ground to a halt, raising questions about whether the world’s biggest civil information technology project will ever be finished”.
He quoted Jon Hoeksma from e-Health Insider who said that
“the key part is stuck”
and added that hospital chief executives did not want to take the system
“until they had seen it put in flawlessly elsewhere”.
The second PAC report, published in January 2009, concluded that the programme’s failures raised questions about the feasibility of the whole project and that the central contracts—the enormous local service provider contracts—were an encumbrance. Only nine months into his job, Matthew Bellamy quit as the chief information officer’s right-hand man. Just before Christmas 2009, the then latest Health Secretary, the right hon. Member for Leigh (Andy Burnham), gave an interview in which he sang the praises of the national programme and said that
“parts of the NHS cannot operate without it”.
Unfortunately for him, the then Chancellor of the Exchequer, the right hon. Member for Edinburgh South West (Mr Darling), took a different view—and said so in a television interview a couple of days later. He said that the national programme was
“not essential for the front line”
and announced that he was imposing a £600 million spending cut that took its budget down from £12.7 billion to a mere £12.1 billion.
Meanwhile, new year 2010 was not a happy one for the iSoft directors. The Financial Services Authority—the chief City regulator—announced that it had laid criminal charges against four former directors of iSoft: Patrick Cryne, the founder and former chairman; Timothy Whiston, the former chief executive; and former directors Stephen Graham and John Whelan. They were accused of conspiracy to make misleading statements. The four denied the charges.
Where are we now? We have yet another NAO report, published on 18 May this year, which states in even more bald terms that
“the aim of creating an electronic record for every NHS patient will not be achieved under the Programme.”
The central aim of the programme will not be achieved under the contract. Several conclusions regrettably emerge about Connecting for Health. The first is about overpaying. It massively overpays: acute trusts are costing £23 million, when they should be about £8 million; the system for mental health and community trusts—RiO—is costing £8.9 million per deployment, when it should be about £1.5 million; and the other systems, such as the picture archiving and communications systems for digital X-rays—PACS—and N3 broadband, which everyone says is not particularly good anyway, are also massively overpriced. I should say in parenthesis that the digital X-rays are very good, but Connecting for Health should not have paid so much for them.
The second conclusion is on de-scoping. Connecting for Health has dealt with the problems it has faced by drastically reducing the scope of what is being delivered, but without corresponding reductions in cost. The third conclusion is the hiding of increased costs. The late deliveries meant there have been no running costs for systems that have not been delivered, and the surplus cash is being used to hide the increasing cost per deployment.
Fourthly, there are serious doubts about the commercial judgment and skill of Connecting for Health. It seems that every contract revision makes things worse. Very little of the originally expected system has been delivered, but despite that, the NHS seems to have little or no commercial cover. The Fujitsu termination, when it was fired from the programme, was farcical and generated massive potential costs and liabilities. The local service providers appear to be running rings around Connecting for Health commercially. As the Financial Times noted on 25 May this year, CSC is offering a one-third reduction in the cost of its contract in return for doing two-thirds less work. As the Cabinet Office observed, that would roughly double the cost compared with the original agreement.
The fifth conclusion is the danger of future high costs. When the contracts finish, there is inadequate provision to manage the systems in future. It takes a special skill to leave trusts stuck with systems that are functionally very poor and out-of-date, which were not deemed adequate nine years ago, and still manage to expose them to enormous future costs over which they will have very little control. That is precisely what Connecting for Health is managing to do. Finally, there is a serious danger that Connecting for Health will put CSC in particular in a monopoly position. The proposed revised agreement may be open to legal challenge from other suppliers who have not had the chance to bid.
What should happen now? It is plain that the NHS IT programme has not worked and there is no evidence that it will work. Rather than squandering another £4 billion to £5 billion, which is still unspent, the NHS should recognise reality. Connecting for Health has failed to achieve its central purpose and should be closed down. I am afraid that it will not help and is now more interested in the preservation of its own position than in protecting the interests of taxpayers. NHS trusts must be set free to choose the systems that meet the needs of patients and medical professionals. They should have the power to source products locally that suit their needs, subject only to common standards.