The £5.8bn question: was NPfIT worth it?

This is presumably what payment by results means. CSC, which did not manage to get Morecambe Bay live with iSoft's Lorenzo until June, received less than half BT's income from NHS Connecting for Health (CfH) in 2009-10, according to figures published by SmartHealthcare.com, despite both firms being contracted to provide similar total services to the National Programme for IT (NPfIT). BT, which continued its work in London and took on new work in the south – at too high a price, according to MP Richard Bacon – gained £470m from CfH in that financial year. Simon Burns' announcement of 9 September, which has ended the highly-centralised version of NPfIT - although not the whole programme - also shows CSC suffering more than BT. The US firm is set to see £500m of cuts to its contracts, although these are yet to be finalised, while BT's haircut is a previously-announced £112m. The financial data just released by the Department of Health serves as an end of term report on NPfIT's financial performance as it changes form. An obvious conclusion may be that, to some extent, CfH has indeed paid suppliers by results. BT has delivered more than CSC, and last year, the agency paid it more. Many criticisms can be made of the programme, but not that it paid suppliers for doing nothing. The question of whether it paid a fair price remains open. The figures also show total spending on NPfIT, as of 31 March this year by both CfH and local NHS trusts, of £5.829bn. For that, the NHS in England has gained some valuable IT infrastructure, but it certainly hasn't fully computerised its records, particularly in hospitals. Where this is happening, it has happened slowly and – on a cost per trust – expensively. Richard Bacon, a member of Parliament's Public Accounts Committee, is calling on the National Audit Office to investigate a particular £546m deal made last year with BT, to support Cerner Millennium in seven trusts in the south of England and install it at three, as well as implementing CSE's RiO at 25 community and mental health trusts. He thinks it should have cost in the region of £100m. By the standards of the original NPfIT deals, perhaps that is realistic. But in a yet-to-be-published interview with SmartHealthcare.com, the former head of BT Health, Sir Jonathan Michael – who is now running Oxford Radcliffe Hospitals, one of the three trusts which will get Cerner under the BT deal – comments that the original NPfIT deals were priced on the basis of identical software implementations at every acute trust. While he was reticent to comment on negotiations at his former employer, he did point out that one would expect a different price if each installation is bespoke – which NPfIT has learnt the hard way is of course the case. Sir Jonathan added that government organisations do themselves no favours by changing specifications for projects after they have been agreed. This will almost certainly be seen as the greatest failing of NPfIT: that it was redesigned massively when already underway. This is a better option than sticking to a failing original plan, but it is also a very expensive one. Whether last year's BT deal was overly rich could be a matter for the NAO, if Richard Bacon has his way. But by last year, the government had burnt too many suppliers with a conceptually flawed scheme. It is not, therefore, a great surprise that what may well have been the only company willing and capable of taking on work in the south struck a very hard bargain. The fault lies ultimately with the politicians and civil servants who got NPfIT wrong at the start. This article was revised on 10 September 2010 in light of the ministerial statement on the future of NPfIT