A group of UK computer science academics have called for the Government to rethink its strategy for reforming the technology infrastructure of the NHS.
In an open letter sent to the Parliamentary Select Committee on Health on Tuesday, 23 academics highlighted their concerns over the changes proposed by the Government in the £6.2bn National Programme for IT (NpfIT). They called for a "thorough, independent technical review" of the scheme.
The academics include some of the best known names in computer science in the UK, such as Professor Ross Anderson, an expert in security, Professor Ewart Carson, a healthcare technology expert at the IEE, and Professor Frank Land, who as a developer of Leo, the world's first business computer, is one of the most distinguished names in UK IT.
The letter points to apparent problems that have emerged at the NPfIT including the issue of whether or not the NPfIT will be delivered within the timescales predicted. The letter also points out that "two of [the] largest suppliers have issued warnings about profit in relation to their work and a third has been fined for inadequate performance".
The two suppliers who admitted difficulties were Accenture, who admitted last week that it expected to make a $450m (£260) loss on its part of the contract, and iSoft, who in January said it expected to earn £55m less than expected from its work with NHS Trusts. BT has been fined for failing to meet pre-set targets.
The project has been beset by other difficulties, and back in November it was reported that its centralised patient booking system — a key component of the strategy — would be delayed.
According to the open letter, the British IT community is united in its misgivings about the all-important strategy. The British Computer Society has added its name to the list of critics, by expressing "concern that NPfIT may show a shortfall of billions of pounds".
The academics cast doubt on just about every aspect of NPfIT, arguing that fundamental issues have not been addressed: "Does NPfIT have a comprehensive, robust technical architecture? Project plan? Detailed design? Have these documents been reviewed by experts of calibre appropriate to the scope of the NPfIT?"
In particular the academics highlight the crucial area of data volumes, questioning whether the Government has taken into account the "volumes of data and traffic that a fully functioning NPfIT will have to support across the thousands of healthcare organisations in England."
Talkback
I have 36 years in frontline IT and, although I'm not 100% up on the NHS Project, I can surmise where it may have gone wrong by stating as briefly as I can how I and my esrtwhile collegaues would have gone about it.
1. There should be a study of the 'business' requirements from end users (GPs, nurses, consultants etc.). The study should not mention any IT terminology. The requirements should be published and agreed with the end users before going any further. The benefits accruing from satisfying these requirements should also be stated. Volumetrics should also be contained in the report. One way of getting these rquirements outside individuals user interviews is a Project Definition Workshop(s).
2. An outline architecture should be drawn up and assessed for 'fit' to the requirements. Alternative archtectures should also be considered, e.g., centralised vs. distributed, operating system based or web-based etc.
These architectures should be assessed by people not involved in the impementation, i.e. independetns, who should aslo carry out a risk assessment at the appropriate point.
3. Once the architecture is agreed, it is possible to start to select componenets to implement the architecture. If available components can't hack the job, it may be necessary to revise the architecture and iterate the exercise. The final components need to reviewed AGAIN against business requirements.
4. A phasing plan needs to be drawn up and not try to implement everything in parallel.
5. The whole shooting match should be reviewed again at this point to check nothing has fallen throught he cracks. We can now start talking about bits and bytes and other wonderful things in IT. Costing should also be reviewed.
6. A pilot and impementation plan should be drawn up, costed and personnel assigned.
7. Other design, implementation and review steps are needed but I will be here all day talking.
On the subject of academics reviewing the project, I feel they are the wrong people - you need people from the trenches.
Also, I can't see what can change now without going back to the drawing board or throwing more money at it. The latter will not work if the thing is inherently not fit for purpose. Everyone with experience in IT knows that the costs of correction rise almost exponentially the further into the project one is.
| don't believe this project has had the full support from the medical profession from the outset.
Also, why is it that most public sector projects always insist on Prince2 and ITIL? Given the dismal public sector IT record, these obviously count for little. As they say 'A foil with a tool is still a fool'.
Happy to ccorrspond with anyone about this (including Richard Granger or Tony Blair).
My 2p for what it is worth.