NHS trust: Not buying through NHS IT saved us £7m A direct tender has enabled Cambridgeshire and Peterborough NHS foundation trust to purchase a patient records system at a cost more than £7m below the price paid by the Department of Health (DH). "We set a capped limit for cost and we introduced this rate at the start of the project," Jane Berenzynskyj, Caerus project manager …
They set a price cap - and surprise surprise, there were some suppliers who agreed to meet it.
Now, it's quite possible that the Trust has done an excellent job, but there are lots of unanswered questions in this story - are the business requirements, functional specs, Ts&Cs and SLAs identical to the more expensive NHS IT version? If not, how do they differ?
Without knowing that, it's impossible to judge the value for money.
You are presuming that the NHS IT version is complete in the first place. I would suggest the Trust are based placed to know what their business needs are, opposed to an overarching bureaucracy. I would expect these elements to be different as each organisation is not a cookie cutter fit for the bloated pig that was CfH.
Of course tenders are submitted with an eye on using overruns to add some fat back, but I still think you're being pretty unfair here. The point is that an independent private IT provider has massively undercut the internal bureaucratic bollocks, and by so much that even with huge overruns and claims of 'mission creep', is still going to be a good deal less.
And the DE are tight-lipped, seemingly unwilling to engage in any kind of defense - which ought to be easy, were they to point out what you just have. Which tells me 'hand, cookie jar, caught in'.
It is nothing to do with tendering or projects creep.
The contract is for ALL NHS to buy in a certain way, the individual PCT or whatever is told where to go from the DH.
The biggest sack of bull ever in the history of anything is the logic and way of working. £35 printers become £150, a £200 dektop becomes £560. And when I worked there £49000 was paid for by the DH on a system that wasn't needed, but had already been paid.
And local IT managers have no power to do anything but adhere because Chief Exec's tell them to, because the SHA tell them..
There is a whole load of brown envelopes passing between MP's, has to be as it makes no sense otherwise. Problem is Cameron will somehow make it look like it was his privatisation methods that somehow saved all this money. Even though GP's are nothing to do with it, the hospitals still have to adhere.
I have heard several times that my GP practice uses different software to everyone else in the area, so that nobody can access their records. My GP practice has, at the same time, told me that the Hospital is dreadfully slow getting patient discharge documents to them.
My father is currently in Hospital A. with an outpatient appointment at Hospital B tomorrow morning. He might have been home by now if the Hospital A doctors had been able to finish a ward round (One emergency, which is fair enough, but that was on the third attempt.)
I have to make a phone call to cancel an appointment.
Which is precisely why (well, one of the zillions of reasons) that NPfIT was unfit for purpose.
The national specification should be about data interchange standards, not software-specific. Then part of the procurement spec for software components and systems is for it to meet this specification.
Simples. And it would have encouraged competition.
Lancashire Teaching Hospitals (www.lthtr.nhs.uk) bought QuadraMed CPR (http://www.quadramed.com/Solutions---Services/Care-Management/QCPR.aspx) an American patient revenue management system. £16 million later it still doesn't work properly (15-20 minutes of work to book a new case appointment), won't store patient records bar some A&E stuff and lab results (a different system had to be bought for records) nor does it talk to any other NHS system automatically.
Oh, and the UI is terrible (http://www.youtube.com/watch?v=UrnOkOXhXQM)
So direct tenders aren't always good.
http://www.youtube.com/watch?v=UrnOkOXhXQM
f**k me sideways.
Hands up who thinks this is a a straight screen scrape from a dumb terminal system?
Who thinks the UI was given to the newbie who's just discovered tabs and drop down menu boxes (but not quite got to the chapter on MsgBox) ?
The word "ULTIMUMPS" is also definitely raising a red flag. MUMPS was one of the grandaddies of computerized patient record systems. Using threaded interpretive techniques (like FORTH) to deliver high capability to large numbers of users on relatively low power hardware (DEC IIRC).
Cutting edge stuff in 1981.
It's command language was slightly more readable than Whitespace.
Having working on NPfIT in the early days this is not unexpected. These trusts seem to think they know better than every other trust and that software that works wonderfully in another hospital couldn't possibly work well in theirs. So we get half-qualified procurement teams saying 'look, shiny!' and buying stuff that doesn't work with anything else and doesn't even meet their needs. I've seen trusts with four different patient records systems becuase a medical consultant on £150k decided one weekend he can create a better system on Access than the hospital already uses.
My other half is an ICU doc and loses (as in dead) tons of patients as a result of systems not being connected, GPs being too busy / lazy / ignorant to respond in good time about past history.
The principle of NPfIT was great - a choice of standardised systems that are inconnected so data was accessible from the other side of the country. It's the execution that was piss-poor. Doctors thinking the data belonged to them, not the patient, and politicians thinking systems that take 18 months to implement OOTB could be customised (to a spec not yet written) and implemented in 4 months in time for the election.
/rant
Granger's lack of a proper spec was a serious problem. Blair's requirement that it be done and dusted in a year was another.
Basically, it amounted to "We have billions for you commercial companies if you can give us this.". Where "This" was something not far off the "Back of a fag packet" as far as specifications go.
None of the providers could agree on what their role was (some were extremely project management heavy, and some were more flexible).
When it came time to implement, it became glaringly obvious that the product just wasn't fit for purpose. Both parties to this have blame to soak up. The board of NPfIT (Granger) and the politicians for not knowing what they wanted and agreeing to fuzzily take on something they weren't skilled enough to know how to work through.. And the commercials for accepting such a fuzzy requirement (hell, I don't take specs that fuzzy when I do bespoke work, and if the client is unsure of something I need to know to implement it properly, I ASK and make sure they KNOW they don't know, so need to think about it before agreeing to do any implementation work).
The commercials should have more than enough experience to know this wasn't a real spec. And without a real spec, you can't do the job and be successful (no common frame of reference).
If they'd done this part properly, the NHS would probably now be a year or two into implementation phase, and it'd work nicely and far more cheaply than the monolithic effort that was created.
The general idea was that the data centres were all built, the infrastructure was provided, and all hospitals would have it at a very low cost per site. The fact that it didn't work for most places early enough, and when things were re-negotiated, knowing the real scope of the work and pricing appropriately, the NHS couldn't afford the cost, meant that the project (well, the central Care Records part anyway; PACS and so on are apparently ticking along very nicely) couldn't continue. Having so few sites implemented meant the cost per hospital was exorbitant.
The savings they are making are probably real in this case, but it's all about scale. I hear that there are consortia now attempting to do the shared Care Records solution of their own back, home grown in the NHS (and just using standardised messages/requests) to exchange necessary info. Good luck to them, as I suspect this may well be the real future of a country wide Care Records Service, done properly, cheaply and efficiently, and working for all people who use it.
Average UK male: 80kg
Average UK female: 65kg
Male:female ratio: 1:1
Ergo approximate 14 patients per ton. Ish. Chop of a couple of limbs, and you'll be back to the exact ton.
Unfortunately, if you want your Soylent Green to be less fatty, we'd need to mix in some low BF Europeans to add some extra texture and improve the protein balance.
When you think about it, the amount of money they're paying to get a system close to what they need but not exact. They could probably hire a team of IT professionals to write a new set of software which works, adheres to all the standards they require, and at the end of it, they'd need to keep on minimal employees to maintain it.
£7 million per hospital for software which mostly works.
Or £7 million for a centrally controlled service, which can then be rolled out to hospitals for free, which adheres to a single universal standard which can then be set so, if a hospital would rather use a third party, that third party would have to adhere to said standard.
@wowfood a system close to what they need but not exact
An interesting and significant phrase. In the commercial world it's been established that it's often a good idea to change your business process to fit an of-the-shelf solution rather than get something tailored to your idiosyncratic requirements. You may even end up with an improved business process as a result - after all, the other users of the system are in comparable businesses. At the very least you get a tested solution that's economical and can be delivered rapidly.
Clearly, health care is not the same as a commercial business, and NHS processes may be very different from processes elsewhere. But a system that is a close fit rather than an exact match could still be the best option.
"Clearly, health care is not the same as a commercial business, and NHS processes may be very different from processes elsewhere."
NHS healthcare != private healthcare, which are commercial businesses.
We use a widely-used off-the-shelf product used by numerous private healthcare firms and (smaller) NHS trusts as well as many firms outside the UK. We've used it as long as computerised record-keeping has really been feasible and expected, and we have evolved with the ever-growing capabilities of the software. As with any software product, there's some creature feep with new versions but we don't buy features we don't need and ignore anything new in the core system which doesn't suit us. New features are almost always requested by the vendor's customers.
As a result, over the past 15 years, our business processes have grown to take into account what we can do with it and many departments (even in more recent years) have got rid of paper record-keeping / appointment-making / etc as the maturity level increases. There's a high satisfaction rating within the organisation with the product; it does what we expect and need.
The Powers That Be, however, have recently decided to invoke change for change's sake (which it often looks like the NHS trusts do, certainly in our PCT areas, and I'm never quite sure why). We don't know what the new product will be, or how much it'll cost, but my bigger worry is that it just won't meet the needs of the business (and our business is patients, don't forget). Whilst I appreciate that business processes can (and will) change, it's never an overnight thing. And there's always the risk that whilst people are getting used to a new way of working, things can get overlooked. And in healthcare, many risks left uncontrolled can end in negligence and even death.
Anonymous, well, because.
"Cambridgeshire and Peterborough's purchase of the patient records system was made as part of its wider electronic clinical information programme, named Caerus after the Greek god of opportunity."
According to Wikipedia...
"Caerus is sometimes considered a daimon spirit rather than a God due to his aggressive nature towards humans. He sought to drink their blood, killing many to do so.
It's the white coat with the stethoscope.
Of f the shelf stuff won't work unless public sector is forced to stop using ridiculous working practices and managers are stopped from building little fiefdoms with procedures as the moat.
Example. I get my eyes checked at a hospital once a year as I'm in a high risk group blah blah blah. NHS have target of everyone being seen within an hour. I'm never actually out of there is less than 6, but to hit the target a junior nurse drags me into a room after 50 minutes and asks me a bunch of pointless questions before sticking me back in the queue for four hours. This crap all has to be logged.
This post has been deleted by its author
And in comparison, my son has had orthodontic appointments at the Bristol dental hospital for the last year - about once every six weeks. They specify a time, wan you that they will lose your appointment if you are more than 10 (I seem to remember) minutes late, and then always see him within 5 minutes (+ and - if we are already there). No failures yet.
There are some improvements that could be made with their booking procedures, but they seem to have their queues well under control with decent estimates of what is fairly routine work.
I used to work for a large organisation who made a number of purchases from govermnent frameworks and "approved" supliers. I've seen standard packages rebadged and purchased at a 500% markup in defiance of advice from in-house experts. I wouldn't be at all surprised if this turned out to be just as good as the NHS reccommended product for a fraction of the price.
I have involved with four major projects with budgets over six figures, each of them had "money saved" by going for a custom solution, each one went to tender, and oddly enough those that knew the budget met it in the tender, but they either over ran contractually (extraneous costs), cost substantially more than expected because of rework, or (worse) cost the business indirectly because it didn't do the job.
This one *might* be different, but a developed, standard, interoperable, tested system that works today, would be a safe bet - expensive, yes, but safe, they know the limits, the capacity, and it's here now, I'll be watching with interest.
I still recall, back in the 90s, coming across an NHS deal for speshul seekrit pricing on PCs. It was a closely guarded secret ... mostly because it turned out the seekrit prices were higher than the very same model PC bought individually on the high street, never mind haggling or going mail-order! Part of the problem seemed to be that someone central actually thought locking in a fixed price for a specific model for three years was a good deal - as in, two years from now, they will still be paying today's prices for today's specification, when everyone else is getting twice the power for less money.
Not much changes, though; last year, another tentacle of government paid £8k for a pair of Amazon EC2 instances - both m1.tiny. Yes, the one you can buy from Amazon for a few hundred a year.
Buying software seems a particular weakness though. Rather than buy something off the shelf and use it, they'll try to specify something custom. That costs a fortune - and of course they never get the spec right, so it costs another fortune to kludge it to do the job it's needed for - or they give up and start again with another pile of cash.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
