NHS IT four years late and over budget The NHS IT project, one of the world's largest public sector IT programmes, is already four years late. Parts of the project are progressing well and some savings have already been seen but Summary Care Records, a key part of the project, are now unlikely to be widely rolled out until 2014 or 2015, rather than the original …
"The scale of the challenge in developing and deploying these systems in the NHS has proved far greater than expected, and the timescales... originally agreed... proved unachievable."
Erm, no. Sorry. No, it didn't.
Any competent and experienced IT professional always knew, and said, that this project was far, far too big, far too complex and completely unfeasible. However, rather that listen to such technical people, they went ahead anyway. Rather like ID cards, and the Cannabis bill. Why bother listening to experts, eh? Its only OUR money that they are pissing up the wall.
1. Commission giant IT contractors on lucrative project. Take huge backhanders.
2. Change the rules so often the project is guaranteed to fail. Ensure the contractors take the blame by promising them future projects if they keep quiet and blacklisting if they point the finger at impossible plans and a civil service that has no interest in making things more efficient.
3. Watch the project fail. Rub hands at prospect of more money going into a giant blackhole and therefore more backhander opportunities.
4. Give press briefings at which anger is declared. Promise big new IT project which will break the old cycle of failure and finally solve the problem. Demand even bigger budget to ensure success.
5. Repeat from step 1.
This entire fiasco is a prime example of how things go wrong when you design software/solutions by committee. When will this goverment (and I say this as a public sector employee myself) learn that designing stuff by committee doesn't work if the people on the committee are: A) Clueless, B) So clueless they don't realise they're clueless, C) Lack the balls to stand up when someone suggests something competely useless, utterly pointless and insanely expensive.
Paris, 'cos she's almost as clueless as the monkeys employed by the Department of Health.
"Erm, no. Sorry. No, it didn't. Any competent and experienced IT professional always knew, and said, that this project was far, far too big, far too complex and completely unfeasible. However, rather that listen to such technical people, they went ahead anyway."
I think you're underestimating the scale of the challenge and the paucity of IT expertise across the NHS - even with a smaller project scope, the project would still be "too big" and at some point the way NHS systems are connected would still need to be standardised - so why not get on with it now rather than later given the suppliers took on all the risk via fixed price contracts?
Many of the connecting legacy NHS systems were poorly defined and documented, with the result that the National Programme has faced an ongoing challenge of integrating hundreds of versions of many different systems - across 30,000 sites. In order to make any progress towards a standardised set of services that are consistent across the NHS, the project would always have had to tackle the complexity and scale of the challenge.
Having worked in Industry (for one of the largest companies in the World), and also worked on the National Programme, my own conclusion is that the root cause of NPfIT's headaches has been the historical lack of standardisation of systems / processes at Trust level.
In most cases Trust IT Directors had (and in many cases still have,) the freedom to make their own systems / technologies decisions with little if any reference to prevailing standards or good practice.
In order to get everyone working at the same level, CFH has had to integrate the legacy and manage upgrades at the same time. In Industry, the historical 'mess' would have been addressed a long time ago - it's akin to letting every store manager of Boots choose their own EPOS systems and then tasking Boots corporate IT to sort out the consequential mess of standards, messages etc.
CFH still has this challenge but cannot dictate what is used locally - some trusts e.g. York, continue to go off and do their own thing. It's very easy to suggest the National Programme should have been smaller / less ambitious- but we'd be talking 20 years to drive standardisation at a time when there's a desperate need to increase the productivity at Trusts level to cater for a rapidly ageing population.
It was recognised early on that this was turned into a poison chalice by Granger, but the salesmen of the consultancies (excepting IBM, Lockheed Martin) couldn't turn away from the siren call). Badly speced, badly managed, bolshie customer - it was a dead duck from before it was let.
Now its 4 years late and we haven't even got to the difficult part yet. I very much doubt it will EVER deliver the real benefits which were wanted from the programme - to break the power of the doctors.
Can it, cancel it, save the money now and don't pay the contractors a penny
damn you, damn you all... I wanted to say that, and that, and that and Matthew... that is the best start to a Friday I've read all week. Nice one.
It is not that mistakes occur, it is not that some projects overrun, it is not that the occasional glitch emerges that requires a rethink... the problem is that as soon as any british agency gets it's greasy little corporate fingers into anything worth more than £5 it becomes a complete and utter shambles. Why is that? How do we maintain this consistency? Is there a prize for the most incompetent use of public money and if not, can I head the Quango to oversee it?
Icon - a good vantage point to watch from and a great place to throw sh1t at, a bit like the public galleries of the Houses of Parliament come to think of it.
There are many components to this system and many companies developing individual modules. Much of it is live, some of it is still being developed.
To classify numerous systems as one project and to say it's late is misleading. Name and shame those providers who are late in delivering subsystems, don't attach the whole project.
1. Participate in highly sophisticated competitive bidding process for new government project.
2. Make sure your bid is by far the lowest. Obviously it's quite impossible to deliver at that cost, but no one will complain. (Except maybe competitors, and they would, wouldn't they?)
3. Work away quietly for a couple of years. Meanwhile, the government blockheads "in charge" will quite certainly ask for 7931 new features, different hardware, and a new type of networking. Smile and agree.
4. Point out that, due to all the changed requirements, project will now cost 20 times as much as originally bid, and take twice as long. If anyone complains, ask politely if they wish the project to become a highly publicised failure which will hang around their individual necks for the rest of their careers. "No? I thought not".
5. Clean up.
6. Repeat ad lib.
In the private sector we have a variation of the rule...
<<Change the rules so often the project is guaranteed to failChange the rules so often the project is guaranteed to fail>>
which basically reads "Change organization so often that no one can be found to take the blame when things go wrong"
"Choose and Book is nearly fully deployed but takeup remains disappointing."
That is hardly surprising. I had to use that system recently. I hate to think of how much money they wasted on implementing that, as from my experience it is utterly useless. Completely pointless from start to finish. It gives you a "choice" which you can only base on artificial statistics which have no meaning and are in no way relevant to your choice. It wastes huge amounts of time, and must add masses of bureaucracy (but to the people trying to cure patients, not to the people running things).
But hey, The Patient Has The Right To Choose!!!!!!
Oh, and it only works on Internet Explorer.
If ever there was a system that could have been designed as greenfield, this was it, the amount of people for whom legacy data was still outstanding decreasing by 2130 a day (due to their deaths.)
This means that without even doing anything at all, old records would become completely unnecessary to migrate at the rate of 700,000 a year.
However, since I know personally many of the "consultants" who've worked in Leeds, and am aware that they wouldn't make a decision that was detrimental to their earning power, this, like many other obvious conclusions will have been disregarded with the "we really should talk a bit more about this, as it's quite promising."
I can only speak from personal experience of the first two or three years of this monumental cockup, but pretty much all of the engineering and technical staff working on this thing new that the project as spec'ed out was unachievable in the original timescale and riddled with problems.
Add to that a bolshie customer (as someone else has already mentioned), a fixed price contract arrangement (which meant that some prime contractors didn't have much opportunity - or the balls - to turn around and stop the customer from constantly requesting changes and moving goalposts) and an excess of rather poor project management and support staff who did very little but cost money and get in the way of real work being done and it's a miracle that it's only four years late and slightly over budget.
I can only assume that they've finally sacked 50% of the project support staff and started having some realistic conversations with the customer. Which I'm sure would come as a pleasant surprise to many people if there's even the remote chance that it's true!
Stop icon, 'cos that's what should have happened to this immense waste of public funds much sooner.
The average margin for large IT company is 4%. This is mainly, in government, because it costs an arm and a leg to bid for and government work and we all have to recover the costs of the 4 out of 5 bids we don't win.
By the way, didn't young Mr. Granger say he was going to screw the IT suppliers into the ground over price on this project. Well that worked didn't it. I expect all the contractors went immediately to to profit recovery plans, which mean, we will deliver exactly what you asked for in the contract, and if you've got it wrong, which we always know you have, we'll screw you into the ground with change control, because what are you going to do re-let the contract, ah you can't, all those lost profit clauses, all that time, and the fact the Minister will have to stand up in the house and say we've failed to deliver.
Well maybe someday the government will learn, it's always cheaper to do your own IT because there is no such thing as a standard government.
As an IT contractor for the Southern sector I can say that the CRS software written by Cerner has been mostly completed, except for some variations from the US version.
What has been holding up the implementation of the CRS system in the South has been the problem of getting the NHS to work to the timescales and be fully commited to work with the local service provider to configure and customise the system.
The NHS takes so long to make decisions and has to have many meetings to decide on configuration, project end dates are extended. The implementation is complex, but even so, it must be possible to install CRS faster if the NHS fully co-operated.
Hopefully the contract reset, when it comes can help in the future.
I work in the public sector in IT Management. I'm responsible for purchasing and projects. However, if I'm over budget, make the wrong purchase decisions or go beyond the quoted timescales for projects, my job is on the line. Why isn't this the case higher up?
Gareth Jones is right there are big wins here. Anyone needing an Xray, CT or MRI in England since last year has seen the benefit from this even if they didn't realise it at the time.
I just don't understand why NPFIT decided to buy into Lorenzo before it even existed. There were plenty of clinical systems out there already capable of doing the job.
I'm thinking that if this project does eventually get finished, it will probably need to be re-programmed and designed anyway. Anything that takes more than a fews years to complete in the I.T. is almost certainly going to be out of date, and use obsolete technology. Its not realistic to have a project that last this long because at some point you're going to have to adapt to changes in technology whilst developing the system.
Not to mention adapt to the continuing changes within the NHS, and most probably the constant spec. changes that come from government sticking their nose in (one would guess, which has killed many a project before).
My point is the fenceposts will have changed after so long, any work done X years ago would certainly be irrelevant today.
The CRS contractor has one fundamental concept entirely wrong.
The reason that millennium takes so long to fit the NHS model is that it is an american system, developed for single sites (Cerner words here, not mine), private healthcare and with ZERO government targets. The terminology is all wrong - even the bloody paper is a different size so standard reports information do not fit.
they can't even spell -appointments 'canceled' has Two (2) L's
'we have NHS numbers, not SS numbers. We have no need of super bill or charging accounts (all present). We have Physcicans, not 'Attending'.
We have clinics, not Ambulatory units' and wards are bloody wards not nurse units. Episodes of care are not encounters and thats just the face side of the win 95 based software with 8 character filenames.
It also has never heard of waiting lists and does not the fit the primary/secondary healthcare referral model.
We are supposed to be on Release 1 by now -they can't even get R0 to function correctly yet.
Staff are trying to use an overly complex - bodged- system against a ticking Government performance clock, with irate patients shouting at them.
To give you some facts from the coalface.
To create an outpatient appointment used to take an average of 2 -4 minutes using a 25 year old Unix based PAS.
To make one now using Choose and Book, non-direct booking and Cerner Millennium takes about 15 minutes -(10 if the operator is a hot shot very rare)
So thats 4 to 6 appointments a Hour.
How many operatives - Hundreds?
Dozens?
No this isn't an insurance company, although the rate of callers is comparable.
The average staff on phones is usually 2-4
and thats an average DCH (district general hospital)
Thats an O/P centre that normally takes about 200-300 calls a day -you do the math
The staff are paid minimum wage and are simply going sick or leaving to work at Tesco's - "Less interesting; but no abuse," as one long time staff very experienced member collected her coat and left the NHS.
On the whole -it would help if the political bullshite stopped and the real insiders told their story
So much for Freedom of Information
@ Simon: "Choose & Book.... only works on Internet Explorer v6" - OM(f)G you better be joking.
@ Andy Gibson: "Why isn't this the case higher up?" - because you don't have 3 hookers, 2 pregnant secretaries and a cocaine habit to maintain.
@ stickman: "MRI in England" - I'm in England; I just got recommended for an MRI scan too. It was done by my Doctor posting a large brown envelope to the local hospital using a pre-printed sticky label and it was VERY efficiently organised - the hospital phoned me the very next day to arrange the procedure.
As the great Omar Khayyam said - He who knows not and knows not he knows not, he is a fool shun him !! Rather describes our civil(??) servants !!
@Giles Jones - What you are saying is akin to that of the water cooling system in the car is not working but, don't worry, the engine is still going great guns !! If whoever that idiot, and I DO mean Idiot, in charge of the entire project cannot see the broader picture by concentrating on too hard on the individual bits, then he should be taken out and made to sit with his bare bottom on an uncooled dual core CPU running at full use !! The smell of roasting nuts will soon convince him of the error of his ways !!
WTF are "Project Managers" for if not to manage the *WHOLE* project. Excuses are unacceptable !!
"According to our CEO, the average margin for [a] large IT company is 4%..."
Oh poor diddums, Tesco made a fortune on that basis. Of course, it's an absurd oversimplification. Pure hardware tends to make about zero percent, but somehow they make it up on volume. Pure software makes about 90 percent gross, and 20-50 percent net depending how big your customer base is and how well your company is run - which is why Bill Gates is so ridiculously rich. Services fluctuate somewhere in between, and depend a lot more on salesmanship and contract engineering. But a move to services tends to be a reliable sign of a company that is headed down the tubes in the near future. Too many skilful contract engineers ready to slip the knife in, and hardly any relationship with skill, experience, and ability.
Since there aren't any big IT companies in the UK (any more) it's academic as far as we're concerned.
Speaking as a doctor who has tried to use the system, I can give you a professional opinion: it's a pile of shit. I have successfully booked a patient into a clinic that the hospital say doesn't exist, and had a patient return to tell me that if I fax the referral over, it'll be much quicker. I can't refer knee problems to the knee expert, only to the orthopaedic department. And then there's the hospital who managed to book a number of heart transplants as day cases - two of which were to be done by an orthopaedic surgeon.
Anonymous, I still have a number of years to go in this job
One of the criticisms in the NAO report was the lack of progress on the Detailed Care Record - but they seem unclear as to whether this is a SSEPR (single shared electronic patient record) - the Lorenzo level 4 model, being one record per patient across all areas - hospital, general practice, community, therapies etc - or a hospital record on the Cerner Millennium model.
All I can say is that with the SSEPR model, there are serious problems in clinical governance (including who can read, who can write, who can correct and how you manage prescribing when a whole host of people are prescribing for the same patient), privacy and confidentiality (how do you define it if all of the county is using the same record, and who is the Data Controller and Caldicott Guardian?) and whether it would be useful and fit for purpose. (I have this vision of an orthopaedic surgeon wading through a GP record.. ;->)
http://shorterlink.co.uk/14325
I find the potential for destruction of general practice records in Lorenzo disturbing - and contract driven reform of the NHS on a 3 single system (one per LSP) basis insane.
As the representative from Fujitsu said at HC2007 "interoperability is not included in the contract".
What was wrong with the evolving interoperability approach we had before NPfIT?
Having just had the 'joy' of helping my in-laws apply for a UK visitor's visa via the government's new online visa4uk.fco.gov.uk site, I'm really not in the least bit surprised the NHS IT project is years late, way over budget, only works in IE6 and overall a bit of a dog's dinner that's past its sell-by date.
Over my (many) years I've seen some utter crap as far as UI goes, but Visa4UK is in a class all on its own. I dread to think what the NHS one is like (if that is it ever fully sees the light of day)...
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
