virtualgryphon.com

News from yesterday indicates that the project to build an integrated healthcare information system for the U.K. is close to collapse. Reaction to a report from the Public Accounts Committee of the House of Commons indicate that key elements of the £13 billion system are not working now, unlikely to work by the projected end of project in 2015, and may never work.

The system, which was supposed to make medical care in the UK “better, safer and faster”, is an end-to-end integrated healthcare information system similar to that currently sought by U.S. healthcare leaders. The U.S. system is one of the projects being targeted for a US$60 billion investment by the Congress and the Obama administration.

The U.K.’s clinical healthcare information technology (CHIT) system evolution has followed the pathway of other, large, expensive IT system failures. CHIT was proposed as a remedy for many of the problematic features of an essential, expensive, and politically sensitive healthcare delivery system. The benefits of CHIT were optimistic extrapolations from demonstrations. The massive project itself was similarly optimistically planned and budgeted. Ironically, controls intended to avoid cost overruns and project delays acted as as incentives for primary contractors to pull out of the project when problems arose.  Progress on relatively easy parts of the system (e.g. incorporation of already working image storage systems, network infrastructure) obscured failure to create and field core functional elements. Fielding of some crippled, partly working components was hailed as “installation” and treated as success despite clear signals that the components made clinical work harder rather than easier. When problems were raised, CHIT proponents classified them as small faults that would be resolved by future software and hardware improvements or even as evidence of “resistance” from clinicians themselves. Over time the goal of making care better, safer, and faster was replaced by the need to simply get the system working irrespective of its impact on care — with a promise to incorporate these features in a future release.

The Commons committee report and the associated interviews and press reports are couched in language suggesting that political support for the system is waning. Significantly, the sunk costs (possibly £4 billion), the lost time and momentum are probably unrecoverable. As with other large scale project failures, the U.K. system owners now face Hobson’s choices. They can abandon the existing system and give up the grand plan of integrated, national CHIT. Or they can persevere with a system so poorly organized and planned that it may cost so much more and take so much more time to fix that its benefits will never equal its costs. Not surprisingly for such a large project, national prestige and macroeconomic factors may have more to do with the choice than any technical assessment. As Nigel Edwards, the Director of Policy for the NHS Confederation observed “There’s a real hazard of doing [with the NPfIT] what we did with Concorde”.

The experience with the U.K. system gives U.S. observers nightmares. One part of the economic stimulus planned by the Obama administration is a huge investment in U.S. CHIT — by some accounts as much as US$ 60 billion.  The claim is that the savings will pay for this investment. If the U.K. experience is any indication, the cost/benefit ratio is not likely to be favorable. More importantly, it is unlikely that the proposed U.S. system can be made to work even for US$ 60 billion.

The U.S. investment will be proportionately somewhat less than the U.K’s. £13 billion project. [U.S. GDP was around US$ 13.8 trillion in 2007 while the U.K.’s was US$ 2.13 trillion giving a GDP ratio of 6.   £13 billion is about US$ 18 billion so the ratio of investments is about 4.6]  Whether integrated CHIT can be created, adopted, and made functional in less time than in the U.K. is unknown. The U.K. system plan encompassed 10 yrs and is, according to the committee report cited above, about 4 yrs behind schedule [The U.K.project was launched in 2002. The project is now producing a behind schedule rate of 6 mo/yr!]

The projected benefit of the integrated U.S. CHIT is supposed to be US$ 77 billion per year, according to a RAND study although a report from the U.S. Congressional Budget Office makes it clear that this is an optimistic estimate.  The RAND report assumed that the U.S. had a functional, national CHIT system and that everyone used it properly.

Experience with other, large-scale IT system failures in the U.S. is not encouraging. The air traffic control system, various military projects, and healthcare system (notably the Kaiser system) IT projects indicate that the investment under consideration in the U.S. is an order of magnitude larger than any IT system previously attempted. Large socio-technical projects do not, to use a computer term, “scale” well.  The present U.S. healthcare system is an exceptionally large enterprise (between US$ 2 and 3 trillion / yr) and it has proven difficult to manage and change. Whether good CHIT can and will be created and whether the net result will yield substantial cost savings are unclear.

References:

BBC News, 2009. Warning over fresh NHS IT delays.

BBC Today, 2009. Tuesday 27 January 2009, 0736. (Audio)  (reported in Tony Collins IT Blog, 2009. Much frustration over NPfIT says BBC correspondent. )

Tony Collins’s IT Blog, 2009. Report of public accounts MPs on NHS IT - main findings.

Tony Collins’s IT Blog, 2009. NHS Confederation: we need NPfIT cash but spent differently.

Tony Collins’s IT Blog, 2009. Much frustration over NPfIT says BBC correspondent.

Tony Collins’s IT Blog, 2009. NPfIT risks heightened says chairman of public accounts MPs.

Tony Collins, 2009. MPs question the future of £12bn NHS IT scheme.

Tony Collins, 2009. Government IT disasters: a clear case for change.

Tony Collins, 2009. NHS IT warnings the government ignored.

Tony Collins’s IT Blog, 2009. NPfIT central contracts will never work - MP tells BBC.

Anna Mathews, 2008.  Savings from Health-Care Computerization May Be Overstated.

Public Accounts Committee, House of Commons, 2009. The National Programme for IT in the NHS: Progress since 2006.

U.S. Congressional Budget Office, 2008.  Evidence on the Costs and Benefits of Health Information Technology.