Abandoned UK NHS IT system has cost £10bn so far | Hacker News Reader

Abandoned UK NHS IT system has cost £10bn so far

theguardian.com

15 pointsdodyg16 comments

As someone with a little first hand experience in the NHS IT (I develop medical imaging software) I saw how at the mercy of the vendors the NHS became in the world of PACS (the systems used to store and retrieve medical images). Not only where they being charged vast amounts but the level of capability delivered was embarrassingly poor. I worked with many clinicians as a university researcher and again when I was working in clinical trials for a pharmaceutical company and they were continually amazed by what we could do with images; I heard the phrase "why can't we do that in the clinic" so often it was one of the reasons I moved into a medical imaging startup. I thought the limitations might be regulatory but after developing and successfully marketing an FDA cleared, CE marked piece of software we came to realise that the biggest obstacle in the UK NHS were the PACS vendors. Essentially they own connectivity to the images and made it almost impossible to integrate into their systems in a way that really addresses the clinical need. It doesn't help that the NHS seems to be filled with people only to eager to encounter a reason NOT to do something (in part I'm sure because everyone is too busy to add more to their plate). The technical competence of the senior folk was also often poor, they were usually senior clinical folk who relied on consultants for advice, often consultants provided by the vendors. The result was it got itself lumbered with long, expensive contracts that tied them into particular vendors which prohibited access to the wider ecosystem of solutions. Thankfully the NHS started to realise this and began rolling down many of these contracts. Hopefully a fresh start this year particularly with the cessation of the problematic Connecting for Health [http://www.connectingforhealth.nhs.uk/] program.

Is there any well run IT system for a comparable national health system anywhere else in the world?

Israel has a single-payer-multiple-competitor system. It's national in the sense that everyone is covered. It is not national in the sense that there are several (5 or 6) competing HMOs, each with their own IT / management / procedures etc.

3 HMOs cover about 85% percent of the population, and of those at least the 3rd and 2nd largest have decent IT. Significant parts of which still run on a mainframe....

I've heard horror stories about the biggest years ago - but they might have upped their game since.

It looks like all around fiasco.

I'm pretty sure anybody with industry experience cringed when the government first launched this project. It strikes me the government would be better off devolving purchasing decisions to a local level where individual units can choose something off the shelf.

I thought the whole idea was to allow sharing of patient data between units - how is that going to work if everyone buys their own system?

e.g. my son has an eye infection at the moment and we've been round 4 different places - GP clinic, optometrist, hospital, eye hospital over the last week. The process of getting the appointments was great (the time from asking for an appointment to the actual appointment time) has never been less than 2 hours - but this caused a problem in that the systems couldn't keep up.

> how is that going to work if everyone buys their own system?

Open standards?

In an ideal world, of course. But these are systems being purchased by the UK public sector....

I'm pretty sure every member of the public with half a brain cringed when the government first launched this project. The woeful reputation earned over the years by (UK) government-led, data-driven IT initiatives speaks for itself.

It would seem that the two primary skills prized by the IT consultants is extremely short term memory and the ability to stick one's fingers in one's ears and go 'la la la - I can't hear you..oh, and here's my bill'.

Devolving the purchasing in itself would not necessarily be a good idea, but I'm all for a six-pronged approach:

1) Develop or adopt robust data handling standards and a certification strategy 2) Develop a common/compliant platform standard and a certification strategy 3) Issue RFQs and proposals to competitive tenderers with the clear understanding that there will be no sole supplier 4) Evaluate and certify the offerings from winning tenderers 5) Allow competitive purchasing decisions at a local level from the certified/approved hardware and software vendors. 6) Implement a rolling auditing, performance re-assessment and certification renewal scheme for all vendors

Time and time again the 'one shoe fits all', single vendor solutions from one of the big consulting organisations have proven to be a disaster, yet back they come when the next future-fiasco is in the planning stages. Cynical me smells something fishy about the whole cosy club.

Your points 1 through 6 sound pretty sensible to me - except that when it came to actually doing any of that it would all be handed across to the very same consulting companies and vendors that are, in my opinion, largely responsible for the mess in the first place.

Quite likely - although any vendor assessment programme would include reference to performance/deliverables in previous projects.

(Actually, if I had my way, certain vendors and consulting groups would be told to fuck off and not bother tendering!)

And that's presumably why you wouldn't be allowed anywhere near making such a decision. You would be trying to get the right thing done (a failing many of us have) whereas the goal of these consulting groups is to make as much money as possible and who cares about delivering anything!

I would love to hear the inside stories on this.

David Craig's book "Plundering the Public Sector" is a bit old, but it gives some details of how these projects were run and for whose benefit (there is a hint in the title):

http://www.amazon.co.uk/Plundering-Public-Sector-David-Craig...

If you can get your hands on it

http://en.wikipedia.org/wiki/Lions,_Donkeys_and_Dinosaurs

written by Lewis Page (Of theregister.co.uk fame) has more stories related to the MoD

About 15 years ago, when this project was just starting, I was a school kid, doing work experience in the IT department of the local hospital. I did stuff like tearing down desktop boxes for repair, and building install images (although I doubt these ever got used, security there was pretty tight.)

One day, everyone in the department trooped over to a lecture theatre, which was full up with all kinds of staff, doctors, support staff etc. It was about the newly conceived NHS IT project, which was at this stage just a patient records system. It was the slickest presentation I'd ever seen, delivered by a similarly slick government guy, who was on tour, preaching the good word about how amazing this new system was going to be. Even to me as a child, it was clear that he was not entirely to be trusted, mainly because he had obviously not made his own powerpoint. I remember a lot of stuff about "preferred partners," and thinking it was all very fishy indeed.

Afterwards, we in IT adjourned to our own meeting room, to exchange views. There was universal despair. Foreheads met table with some force. The meeting basically consisted of going round the circle, each battle-hardened techie sharing their particular view of why this would never work. "I once built a database of every vauxhall car in europe, with all their service records centralised. That was a fraction of the size of this, and it was still a fucking nightmare, even though the people designing it and the people building it were the same people, and we had absolute technological discretion." And so on.

The pretty much universal consensus was that this would be a disaster, a plague on all their lives for decades to come, for some, the rest of their careers. I credit this moment with warning me off corporate computing forever. Now I write ruby in a place where I am not obligated to wear shoes. Thanks NHS!

EDIT: as a part of my time there I also went to the records department, which was like a giant library with shelves up to the ceiling and moving ladders. On these shelves were the paper patient files, big fat green folders full of assorted documents bound together with those treasury tag things. This was what the slick guy wanted Fujitsu and friends to replace.

I saw the little green records request slips that came into the archive department. Doctors wrote them out when they wanted someone's file. If the file had already been given out to someone else, the archivist would find a slip of paper in it's place on the shelf, detailing it's whereabouts. They would cross reference various numbers, and then send the request slip on to the noted location of the file. When it got there it was quite possible that the file would no longer be at that location, having been given onwards to someone else. Doctors were under no obligation to return the files when they were finished with them, they could give them directly to the next person who needed them. However, they would have to write and hold on to a slip detailing where it had gone. In this way, the little green slips formed a linked list, sometimes 5 levels deep. There was much beeping of barcode scanners: they had built an in house computerised system for tracking the locations of the paper files.

I asked, and apparently the files never got lost. The archivists enforced this regime with a fascistic level of seriousness and intensity. Any doctor who didn't fill out the green forms correctly would face their wrath, and they didn't do this because they knew that if the system wasn't adhered to then they wouldn't be able to get the patient records they needed. The doctors had built the time-delay for file retrieval into their procedures, and everything ran smoothly around it.

This taught me ultimate respect for paper, and the folly of just throwing away a system which highly skilled people have built up over decades, even if you think your whizz-bang technology will be better.

Reply on news.ycombinator.com