Saturday, August 01, 2009

Scrapping The NHS Supercomputer


Best place for it

Earlier in the week, we said that Mr Cam's pledge to scrap the notorious NHS Supercomputer (the National Programme for IT - NPfIT) would save next to nothing. Having taken a closer look at the figures, we now think that we - yes, we, not Mr Cam - were wrong.

We now think scrapping NPfIT might save us around £1bn pa over the next five years. And that's not to be sniffed at.

We've blogged the Supercomputer many times over the years (see collection here). Foisted on NHS clinicians against their wishes, it is now budgeted at over five times its original supposed £2.3bn cost, and is running years behind schedule. Chuck in serious and unresolved concerns about patient confidentiality, and you are looking at a Grade A unmitigated disaster (one that was personally ordered by Bliar at a No 10 summit in 2002).

As always with such disasters, reliable up-to-date cost figures are hard to come by. But we do have a Public Accounts Committee report from earlier this year, and an associated National Audit Office report from 2008.

In summary, they give us a total cost estimate as at March 2008 of £12.7bn. That includes £6.8bn on the "core contracts" with IT suppliers, a further £0.7bn on later contract additions (scope creep), and £3.6bn on local implementation costs to be met by NHS trusts.


Unfortunately, those figures are at 2004-05 prices, so we need to adjust them. The IT contracts are effectively indexed against the RPI, which has already risen by about 15% since April 2004. And although we do not have a year-to-year spend profile, given that we are now only halfway through the supposed final delivery timetable, we can quite reasonably inflate the total by 15%. Which gives a total estimated budget spend in cash terms of around £14.6bn.

As at March 2008, just £3.6bn of that had been spent (c 25%), including £2.3bn on IT contractors. Which means there was still about £11bn left to spend over the next 7 years. And of that, £6.3bn will go to the contractors (= £7.5bn times 1.15 minus £2.3bn), with £4.7bn to be spent within the NHS, either centrally or locally.

Let's assume that the £11bn is spread evenly over the years from 2008-09 to 2015-16 (the end-date). That's a conservative assumption because in reality the spend is running well behind schedule and future inflation will lift the later year spends. But to be conservative, let's assume it anyway.

So that means future spending up to 2015-16 is c £1.6bn pa, of which the contractors wil be getting roughly £0.9bn pa.

The question is how much of that we could escape by scrapping the project now?

As per, there is a total lack of info on this. But we do know that the Dept of Health has signed contracts with the IT suppliers that commit the NHS to make payments whether or not the individual health trusts use the new systems or not. Which means that the contractors have quite sensibly locked us in: cancellation would incur financial penalties.

But how big are those penalties?

Unfortunately, we don't know - the contracts have not been published.

But bearing in mind that these contractors are struggling to deliver the project anyway (two of them - Accenture and Fujitsu having already dropped out), and bearing in mind that they will want to stay on the right side of Mr Cam, there ought to be an opportunity to negotiate a fairly keen exit deal.

It's a shot in the dark, but let's say we could get out for a 30% penalty charge (which is pretty hefty) spread over the remainder of the contract life. Then we'd save c £0.6bn pa on the contractors' costs.

To that, we could add savings within the NHS, both centrally and locally, of up to £0.7bn pa. True, some of those savings might need to be spent on alternative systems, but we should remember that according to the DoH, the Supercomputer was never meant to cut NHS costs. Total savings are estimated by the DoH at only £119m pa, so by the same token, scrapping it ought not to increase costs by much.

Overall, we estimate that scrapping the Supercomputer - net of penalty cancellation charges - could save us around £1.2bn pa (= £0.6bn saving on contractors' costs plus £0.7bn saving on internal NHS costs minus £0.1bn offsetting additional NHS costs).

To be ultra-conservative let's call it £1bn pa.

Now, can anyone see anything we've missed?

Surely even IT contractors wouldn't demand more than a 30% (=£1.5bn over the remaining five years) penalty cancellation charge? Not if they want to be seen as patriots, anyway.

PS In case anyone thinks we should press on with the project because we've already spent too much to stop, just read the damning PAC report. The guts of the new system - the patient care record - is nowhere near finished, and what has been delivered has been found seriously wanting by users. It's finally time to pull the plug.

No comments:

Post a Comment