“Stuff goes wrong all the time. You know, computers do fail. But what we’re seeing is a sort of hysterical coverage. What I should be judged on is whether we’re fixing it quickly and ensuring it’s as good as anything else anywhere on the planet. Measure me on those things and I know we will not be found lacking.”

For a man better known for savaging suppliers, with an apparent ‘lead me, follow me, or get out of my way’ attitude, 42-year-old Richard Granger, director general of IT, NHS, is surprisingly plaintive. We met in Whitehall a few weeks before he announced his departure at the end of this year after five years in what must be the biggest, highest profile civilian CIO job in Europe.

"“There is a little coterie of people out there who are alleged experts and who worked on this programme. They were dismissed for reasons of non-performance or in one case, for breach of commercial confidentiality”"

Richard Granger, director general of IT, NHS

Granger has heard a lot of domestic condemnation of his role over the past five years. Critics argue the project is too complex; that it should have a more localised approach; or one based on smartcards; and that his mismanagement of finances has resulted in a £12 billion bill that is fast rocketing to £50bn. Or just pick a number.

Granger could never win at the NHS but he brought a vigour to the role, which alienated many; the contractors may be the same old, same old but they are being managed and paid under rigorous operational delivery targets; and Granger is clearly as fanatical about value for the taxpayer as he is passionate about the NHS.

Granger has been lambasted for budget overruns but, “it’s complete garbage,” he says.

Ditto with scope. This project is like Topsy. It just keeps growing and there is no evidence that demand will decrease. In fact, all evidence points to the contrary. Granger, having announced his resignation after five years in the job, was ‘accused’ of leaving the project not even half-finished. The fact is, NHS IT will never be done. And five years in the job is longer than most CIOs stay in one position. The average tenure is just over three years. “The scope is huge and everybody had problems with timescales. Accenture had terrible difficulties but the original Strategy Review in 2002 [just before he joined the NHS] said we could do most of it by the backend of 2006/2007,” he says.

Back then it was it was a three-year programme and planned to cost around £5bn. That was followed by the 2002 Spending Review.

“They pitched for £2.5bn and I’ll never know the relationship between the original estimate of £5bn and the revised £2.5bn because there were no working calculations for any of it, as far as I can see,” he says. The project was then allocated £3bn over three years. Granger believes that part of the deep confusion over money is that while the strategy document set out a work programme that would last several years, public sector funding tends to be done in three-year cycles.

“There was confusion about what it would cost to have enterprise systems built and running on a permanent basis in the NHS. That was one question but the other question was, ‘how much money will you spend in the forthcoming spending round?’” He contends fiercely: “I’m under spent, not over budget and one reason I am under spent is because some of my programme [mainly the core spine and electronic care records] is running late. All sorts of wild numbers are being produced, the best one being £50bn from Tory MP Richard Bacon. At the end of March 2007 we had actually spent £1.5bn,” he says. Disbelievers, he suggests, should check the audited accounts of the public companies that supply the NHS, which show no sign of inflated NHS spend. “This isn’t complicated algebra.”

Project creep and dodgy numbers

NHS financing will haunt Granger for the rest of his career and it is unlikely the final cost will ever be apparent. The National Programme today is said to be running at around £12bn but Granger takes issue with the current calculations.

"“Sometimes we put stuff in that I’m just ashamed of. Some of the stuff that Cerner has put in recently is appalling... Now they’re being held to account because that’s my job”"

Richard Granger, director general of IT, NHS

Last year’s first big ticket, headline-grabbing number was £6.2bn, which he insists is the value of the contracts, as opposed to him going way over budget. “It’s difficult to knock that down but they are completely different things. One number was an estimate, with no engineering basis, for how much money would get spent in the first three years of the programme to build the asset. The other is a contract value based on a private sector estimate of the build and operate costs over a 10-year period. They are entirely different. We are not over budget on the original estimate.”

Also, Granger says the scope of the current programme is far larger than the original plan, something with which few commentators would disagree. It is a clear and fundamental problem with NHS IT delivery.

“One thing completely missing from the 2002 Strategy was picture archiving. Subsequently, many doctors and patient groups lobbied me for that, saying it was a big omission. So I went to the Treasury and they gave me another £1bn to do it,” he says. “I really don’t know which school of budget management some of these people come from but that is not going over budget. That is a change in scope. Then there were other things demanded by the user base not included in the original scope. The Quality Management and Analysis System (QMAS), the GP payments system for example – and I ended up taking over the NHS mail contract and retendering it.”

He argues there are two levels of confusion. One, there are a number of existing systems and those budgets have become enmeshed in his. Two, the rescoping and subsequent perception of engorgement. “The email systems service 250,000 people at 24p/user/month, also to SME and fax. That’s eating about £10m/year. It’s excellent value to the taxpayer but wasn’t in my budget originally.” In short: “No cost overrun, no financial mismanagement and we are not paying suppliers a penny more than we said we were about to.”

Granger says the muddled economics were complicated further when another question was posed: “What is the total cost of ownership of these systems over 10 years?’ and that’s a very different question, again. These costs, for example, would include data conversion, training new users, retraining and end user hardware upgrades. The response from the Department of Health (DoH) was ‘five to six times the contract value’, which is not a silly number but then the headlines were about the National Programme being up to 10 times the original budget. The next phase of costing saw some interesting exchanges between members of the National Audit Office (NAO) and Public Accounts Committee (PAC) but by then, the mythology had started to be cemented into people’s minds.” That NAO report last summer still rankles with Granger.

“They went round and added up all the costs they could find and singled me out for special treatment. Cost is ordinarily defined in this environment with benefit having been subtracted. That is, net cost; net benefit. We just had costs. If someone went on a training course, that got added in as the cost of a system. That’s when they came up with the £12bn number. Under significant pressure, I and the DoH, decided to agree to an NAO report that said the total cost of the NHS Programme would be in the order of £12bn. Notwithstanding that, the costs of the contracts that I manage have not overrun. They are under spent, currently by the best part of a billion.”

Granger on suppliers

“One of the things I’ve been careful not to do in the NHS is create the supplier Stockhausen syndrome.” That is, when things go pear shaped, the CIO goes native, sticking slavishly to an agreed line with the supplier, covering asses and, usually, paying extra for the privilege.

“One supplier asked for an extra £500m to deal with cost overruns. He received a succinct refusal but there are many places where the response would have been different; where threats of bad publicity and contract disputes would persuade an organisation to start bunging millions of pounds a month in addition to the existing contract, just to cover up,” says Granger.

A key requirement of any CIO is that he chooses suppliers likely to be around for the long-term. The Accenture/iSoft combination has been seen as a failure of supplier selection on Granger’s part. Not so, he says.

“Last year iSoft was capitalised at around £500m so who’d have thought they’d be worth only 10 per cent of that a year later. [When assessing potential partners] we check the installed base; the development shop; balance sheet strength and the capability to build new products… and I never did any business with iSoft.

“I don’t have a contract with iSoft, I have a contract with CSC. I didn’t do business with iSoft because I didn’t think they were big enough to act as a prime contractor, and I was right. I thought exactly the same of Cerner. Its market cap four years ago was not dissimilar to iSoft, bigger but not twice as big or twice as healthy.” He says he looks to CSC as a prime contractor to address any difficulties: “There was a known risk of an ISV lacking capitalisation to build new product.”

And his take on Accenture? “When they work with a mature, high quality vendor that recognises Accenture as in charge and they’re doing it their way, you get a quite good deal and they’ll do a good job. As with the picture archiving delivery.”

And the Isoft/Accenture/CSC debacle? “Well, neither iSoft nor Accenture was willing to recognise its dependency on the other. But it’s a co-dependent situation. As the IT industry matures, prime contractors will get into delivery programmes with big and small scale organisations but they need to realise it’s a marriage and if they have a bust up, they won’t deliver. We were able to negate the contract with no additional expenditure to the taxpayer. Did it cost taxpayers money to change contractors? No. We fought hard to protect taxpayers’ interests, to stay focused on getting what we paid for and only paying for what we got. I’d love to meet other CIOs who run a commercial negotiation like that with Accenture. When the news broke, three CIOs rang me to ask how on earth we’d managed it.”

He says Accenture’s “work in progress far exceeded the value of the asset we received and they had a significant write off. It was right that they bore that write off. It might be a unique situation in this industry and may cause me some personal difficulties in the future but it was the right thing.”

Prejudicial to public sector success

“One thing I’m very proud of is we’ve not had a large scale system build out that’s been a failure for which the taxpayer has had to pick up the tab.”

That’s not the same as saying there have not been high profile system failures but Granger believes people are trying to create a perception that NHS IT is a disaster and should be written off. “But where are the large systems, the £10m, £100m, £200m software builds that got quietly shut down, a few people got fired and it all got covered up? They don’t exist here. There have been some timeframe and normal lifecycle difficulties but they’ve been blown out of proportion.

“Take the Greater Manchester Patient Administration system. In the same articles we read that none of this was working yet duplicate records are being created. Well, which one was it? It can’t be both. Not working or creating 400 duplicate records that had to be deleted manually until we had a software fix. Is that really a big deal? I don’t think it’s good, it needs managing but didn’t a whole BACS run fail one night not that long ago? How many millions of cheques didn’t get processed. There is a huge expectation of us but at the same time public sector IT projects are somehow expected to fail. We need to be more mature about the disruption that we’re going to go through to get to that other place. We have a hobbyhorse in this country about public sector IT. Some people make a career out of it and it shows a shameful lack of objectivity. I’d like to judged on the same standards as everyone else.”


“Either people are really stupid or evil. It’s difficult to be compassionate with people who claim that suppliers are going out of business because they’re not getting paid or they were withdrawing from wishing to do services within the NHS. At the same time, they are saying they [the suppliers] have been bunged millions of pounds that weren’t budgeted for. It’s wicked or stupid.”

Granger accepts that commentators, media and general public have little recourse but to accept the findings of the PAC and the NAO. If the public account bodies cannot make sense of the NHS numbers, then who can

“Yes, I know. It then becomes public record. What is even more insidious is that on a large programme I’ve been running for nearly five years, in an organisation that failed to do this job in the past, repeatedly – they tried in the early 90s and again in the late 90s, so this is their third shot – there were a number of people around here who had been involved in those failed attempts.

“When I turned up here the backend of 2002, they were pretty annoyed because I knew all the call systems from congestion charging. Should I have kept people who were palpably incompetent and who are now producing negative coverage for me?

“Should I have saved the taxpayer money or taken the easier path by keeping them quiet on the payroll?

“There is a little coterie of people out there who are alleged experts and who worked on this programme. They were dismissed for reasons of non-performance or in one case, for breach of commercial confidentiality.

“He actually sent our financial model to a supplier and that’s why we suspended him. He then resigned which is an answer in itself.

“Who has contributed evidence to public accounts committees? For just about every figure quoted as expert in this programme, I’ve got HR files on them. They generate a piece of opinion and that opinion substantiates their world view.

Shades of Stockhausen syndrome perhaps but suffice to say: "They are no longer within this tent but outside it making it wet."


“When we wrote the contracts, it was very clear we couldn’t build and deliver this stuff in three years. It would take longer unless we wanted to wear the completion and operate risk in the public sector. I didn’t want to do that because I didn’t think it was a sensible position. I didn’t want to build a large public sector IT employee, asset and delivery shop.

“There was strong evidence to the contrary every time a large public sector organisation was market tested. The public sector outsourcing deluge of the 1990s illustrated that it would not stand on a make or buy valuation and it would be far better to get the private sector, with its assets and market leverage, to operate it.

“I knew from other public sector jobs I’ve worked on that I didn’t want the public sector to bear the completion risk of the construction of bespoke software. That’s a very uncomfortable place to be if you’re a public servant – although I’m equally sure it’s not a comfortable place to be if you’re in the private sector and doing it for me.

“Also, I didn’t want to bear the costs for the packaged software as well, which is easier but still not without risk. So we put together deals that balanced out the construction and completion risks with the lower risk of an operating period.”

Hence the construction of 10-year operating periods in return for the build period. “Those contracts provide the funding for that asset to be operated once it’s fit for service. There’s no point me building a network, paying a commissioning fee and then saying, sorry guys, I’ve no money to run it.”

So, the network contract with BT, for example, saw the supplier awarded a sum that would be quadrupled over the course of next seven years. “I’d like to do more contracts like this, with an output-based not technical specification, that are a balance between micro and informed management. They are based on the congestion charge programme, which has been commended.

“We used almost exactly the same term sheet. It’s the same structure, same lawyers, same negotiating conditions and same contractor management arrangements.

“I find it absolutely astounding to read articles comparing it unfavourably with the congestion systems. A large part of the same team – an elite team – took pay cuts to join the NHS and work on this. There is a genealogy of expertise here around financing risk transfer.”


“I sit in large programmes leading teams of people. If it goes wrong it’s at the corporate contractual interface, we rarely have fundamental equipment failure. We have the support to drive through changes and political support has never wavered.”

It would be reasonable to surmise there is a connection between Tony Blair’s long goodbye and the announcement of Granger’s departure at the end of the year. But in the field, Granger has been accused of riding roughshod over the requirements of the user base and local needs

“The distribution of cost/benefit is a difficult part of seducing people to take new systems. GPs for example, bore the brunt of collecting addresses but hospitals are more likely to see the benefits.”

Does that mean he thinks there should be a mandate for user buy-in with public sector contracts

“Our users are highly educated. They often have quite strong opinions. I think you can do something like that in a clerical environment, like the computerisation of social security and quite a lot of my team, including me, worked on that. You built a system, test it in some offices and then roll it out. If you were a clerk administering income support, that was the system you used. There was no alternative

He says the NHS is a far more complicated situation. “We were not going from a clerical process to a computerised process in a nationally controlled organisation. These organisations are statutory independent bodies, especially if they are foundation trusts. They buy services from the private sector and they’ve been investing a billion pounds per year in computers. Two years ago that was money for at least a decade, if not longer. You’ve got a lot of existing assets you must sweat

“We get a lot of views from the end user community about what is right and what is wrong and we must have a mixture of products that hopefully makes their lives easier, although sometimes we fail to do that miserably.

He adds: “Sometimes we put stuff in that I’m just ashamed of. Some of the stuff that Cerner has put in recently is appalling. It really isn’t usable because they have been building a system with Fujitsu without listening to what the end users want. They have taken some account but they then had to take a lot more. Now they’re being held to account because that’s my job.”