FrontDoor24: A project that has failed and is "Slowly" becoming one of the biggest IT disasters since Marconi and the early warning computer system that never was after £350 spent in the 70's . The NHS IT system is dead in the water. The man at the top Richard "the buck stops here" Granger has made a series of miscalculations that will cost the UK £Billions no doubt he will be rewarded for this monumental F** up by a night hood and massive leaving bonus.

Fujitsu became the second major supplier to withdraw from the £12bn National Programme for NHS IT yesterday, prompting another round of the perennial game of "why do government technology projects always go wrong?"

Four years since it signed the original £896m deal for upgrading hospital systems to support a nationwide electronic bookings and digital patient records, the Japanese company could not reach an agreement with Connecting for Health (CfH), the agency running the programme, about how changes to the requirements should be reflected in the payment terms. "Regrettably and despite best efforts by all parties, it has not been possible to reach an agreement on the core Fujitsu contract that is acceptable to all parties," CfH said.

The history of public sector IT is chequered, at best. In the 1990s, the first outsourcing fad led to a run of high-profile disasters, such as queues round the block at the Passport Agency. The nadir was Libra, the magistrates' courts scheme – which ended up costing £390m, three times its original budget, and was described by the Public Accounts Committee as "one of the worst private finance initiative deals that we have seen".

The public sector, and the taxpayer, thought the IT industry was over-selling its capabilities with dollar signs in its eyes. The Government did not get off unscathed either. Technology suppliers muttered about the lack of management capability in Whitehall, about the nightmares of working for a client that never knew what it wanted, and about the impossibility of delivering programmes reliant on seismic procedural changes without proper support from the very top of the organisation.

But the NHS programme was supposed to be different. Richard Granger, a former partner at Deloitte who had been responsible for London's Congestion Charge scheme, was appointed to run NPfIT in 2002 and took a famously hard line. The contract terms were so onerous that one bidder, Lockheed Martin, walked away during the procurement, and once the deals were in place Mr Granger himself said he would "hold suppliers' feet to the fire until the smell of burning flesh is overpowering".

The key difference with Mr Granger's contracts was that suppliers were only paid upon delivery. But delays – some down to the suppliers, some not – caused cash flow problems. Accenture, which took two of the regional implementation deals, worth a total of £1.97bn, walked away in 2006 because it could not see a way for the programme to pay. Fujitsu took longer but reached the same conclusion.

Part of the problem is politics. Part of it is change. Since the contracts were signed five years ago, the NHS has undergone a massive re-organisation: 28 Strategic Health Authorities were created in 2002, and in 2006 they were then merged into 10. Tony Blair provided high-profile backing, creating a hope that the various factions, clinical and administrative, could be corralled. But Gordon Brown's priorities are seen to be elsewhere. The result is a different world from the one in which the contracts and the technical specification were set.

It is a recurrent problem for public-sector technology programmes. "What needs to be looked at is how these contracts evolve over time," Eric Woods, the government practice director at Ovum, said. "They need to be able to take in the emerging requirements that are inevitable in long term projects."

There has been progress. Recent re-negotiations of outsourcing arrangement at HM Revenue and Customs and the Department for Work and Pensions resulted in contracts that were longer, but paid less well. But few of the glut of IT investments in the past half decade have been entirely successful. Problems with the new, £100m personnel and payment system caused outrage in the armed forces last year; the £2bn criminal justice scheme made some headway but fell short of its all-singing, all-dancing promises; the Rural Payments Agency fiasco saw thousands of farmers underpaid; the tax credits debacle saw more than a million claims unprocessed; the Child Support Agency was scrapped altogether.

The one big scheme left on the books is identity cards. The biggest lesson to be learned from the NHS is not about contract terms or technical requirements, but is about politics. ID cards are as much of a political hot potato as the NHS programme was, and we are heading for a general election. Unless the practicalities of building complex systems can be squared with the even more complex organisations behind them, the prognosis is not good.

"We have not yet got right how major public-sector initiatives requiring significant IT investment need to be managed and interwoven with the requirements of policy makers," Mr Woods said. "When the wheels come off in terms of the budgets and the timescales, the major problem is to reconcile those two sides of the programme."