The concerns of the NHS23, the academics who wrote an open letter calling for an independent review of NPfIT, are not yet allayed. Tony Solomonides, one of the 23, spoke at the BCSHIF meeting on 22 January. Helen Boddy reports.

When the National Programme for IT (NPfIT) was first announced Tony said he had a sense of relief. He had been working on a survey of independent databases for Information for Health - and the Local Implementation Strategy Committee - which was just not adding up. They had found 300 different databases existed, 30 were useful but they thought they could only pay for three.

'The NPfIT announcement made me think "phew! - the problem is being taken out of my hands,"' he said. 'The NHS23 have been painted as anti-NPfIT, but we are not; we're highly committed to sound healthcare information systems. We would still be in favour of the main goals of the programme, and we recognise that some things have changed, but there is still a need for a review.'

Nevertheless, by April 2006, Tony argued that both public and private signs showed that things were going wrong with NPfIT, and the 23 academics signed an open letter to the House of Commons health select committee to call for an independent review.

Several among the 23 have been very active, stressed Tony, and they were from different disciplines - computer science, software engineering, and information systems. Tony identified himself with both computer science and information systems. He started out in clinical informatics and moved into health informatics in 1998.

Following their letter, five to six of the NHS23 were invited to talk to Richard Granger, the then head of Connecting for Health (CFH). Although Tony was not among the group, what he heard about the meeting reinforced the concerns in their letter. The most worrying point for him was that there was a 'double blind' architectural approach attributed to commercial confidentiality.

The concerns of the NHS23 have continued. One worrying sign was that individuals who expressed doubts or objections, and in particular one local service provider employee who developed a peer critique of NPfIT, were disowned and silenced.

'There was a feeling that anyone who stepped out of line would be silenced,' said Tony.

Although NPfIT is now moving to local implementation, as announced late last year, Tony thinks it is too little too late. 'The lid was on hard, and has now been taken off a bit,' he said. Local teams have responsibility but they have little choice. What if they don't know where to turn? Will they just do nothing?'

Some of the NHS23 continue to be active in commenting and making critiques on NPfIT. In a recent discussion on BBC TV's Newsnight about government data loss, Ross Anderson talked about how the lessons learned should be applied to the development of electronic health records.

'There are particular concerns about role-based access controls,' said Tony. 'CFH's own analysis shows there are 40 million possible combinations of access controls. Therefore there will need to be a compromise, and some data will be put at risk.'

Martyn Thomas - in his evidence to the Healthcare Select Committee - offered examples from DERA's review of the Swanwick En-Route Air Traffic Centre's software. This was conducted while the centre remained in operation, and, Tony, suggested, could help in devising a review for NPfIT.

Closely related to Martyn's critique is that of Brian Randell, based on the issues of centralisation, evolutionary acquisition, socio-technical systems, and constructive reviews.  These themes were developed in a different talk to the BCS Socio-Technical Systems Specialist Group on 7 February 2008.

Other critiques have come from the information systems point of view - stressing organisational issues, change management, unintended and counter-intuitive consequences. These have mainly been by Colin Tully, Frank Land and Mike Smith and to some extent Ray Ison, James Backhouse and Tony Solomonides.

Colin Tully was remembered as saying: 'How can observing yourself and your problems, and seeing what works, and what doesn't, be bad?'

Ongoing actions include Tony supervising a student, Mark Olive, at the University of West England, who is studying evidence from practice. He is looking at integrated care pathways and variance, knowledge management versus control of the clinician, which is seen by some as the Taylorisation of medicine.

Meanwhile, Harold Thimbleby has developed a twin track critique, on the one hand of the usability of the systems and the other in conjunction with the Welsh programme.

In summary, Tony said that the NHS23 remain critical of:

• relations with stakeholders, poor requirements process;
• the architectural approach, especially:
  - dependability/fault tolerance;
  - security;
  - centralised storage.
• slow, late and unreflecting change in strategy.

Tony was asked what he thought could now be done, given that the programme is in full swing. He suggested that Martyn Thomas could advise the government on how to conduct a review that would not stop the programme while underway. Examples could be drawn from the MoD. A review could also clear up how the local process should work.

An audience member from CFH countered that implementation is now calming down and working better locally.

A website containing over 350 pages of critical comment on NPfIT can be found at: www.nhs-it.info

A tribute to Colin Tully

EurIng Professor Colin Tully MA (Cantab) FBCS CITP CEng, a leading light in the NHS23, passed away on 27 December at the age of 71.

Tony Solomonides began his presentation with a tribute to Colin, whom he particularly remembered for his astute questioning in his role of external examiner and more recently for his incisive analysis of the National Programme for IT.

The BCS Health Informatics Forum (BCSHIF) added their condolences. Colin had a long association with BCS and the forum. He played an active part in Healthcare Computing conferences and the Primary Health Care Specialist Group.

Colin will probably be best remembered for his research into the domains of software engineering and information systems, including software processes, capability maturity models, software life cycles and method and tool integration.

He retired from the post of associate dean at Middlesex last year, but was still very much part of the academic community.

March 2008