Many of us negotiating the original National Programme for IT (NPfIT) contracts questioned whether such an ambitious project could possibly work. The NAO has again made clear that we were right to be concerned, writes John Enstone, lawyer with Faegre & Benson LLP.

The lethal cocktail of heroic specifications, a preference for undeveloped systems over already proven ones, a focus on draconian terms applied to bidders with deep pockets willing and able to endure punishment, the near exclusion of the user community, and frequent missives from Downing Street heralded the victory of a political agenda over the brutal realities of large and complex IT projects. Bidders, for their part, saw pound signs galore and reputational advantages, so accepted the foregoing and reduced contingencies and margins in order to win.

Criticisms are easy, solutions appreciably more difficult. The recent NAO report offers a good starting point.

Two themes recur with worrying frequency. First, the deliverables seem to be a moving target, which alone is enough to drive up costs and delay implementation. Secondly, the department and the suppliers have adopted the protective behaviours of those seeking to avoid blame in the face of possible defaults and legal claims.

Steps to success

What the NAO does not seem to say is also important, namely that the new care records system (as a piece of technology) is somehow fatally flawed. So, assuming that technically the system is or can be made acceptable, it is time to consider a few fundamental, remedial steps to successful implementation. Create an environment in which the players are comfortable talking about compromises and cooperation. For all the legal remedies that the contracts undoubtedly contain, they are not going to deliver a working system. A "without prejudice" approach can suspend the cycle of legally strategic, defensive assertions that players in trouble are prone to make.

Stipulate a short period for discussions and decision-making and treat time as being of the essence. Everyone has had long enough to work out what they want and can do, so there is little need for another lengthy round of submissions. Focus on only a handful of requirements from each side. Ask the suppliers what they can realistically deliver on the basis of existing (or soon to be available) functionality in each of the next four years, and how much it will cost.

Ask the department and the users what they can accept, and then negotiate. Freeze the deliverables, timetable and price - change orders are anathema to achieving a timely, cost-effective solution. Remember that it is much easier to adapt existing data and the like to a system than the reverse. Suspend the remedies that the contracting parties have for breach provided they live up to what they have said they can deliver and accept.

No quick-fix

There are no simple solutions to fixing complex projects, but there are well-tested principles that work.

With a new(ish) government and some pragmatism on all sides there is hope that a system that the NHS very much needs can be delivered. One thing is certain, however: what the players have done so far is not the answer.

This article was first published online by Computer Weekly on 23 May 2011.