Published: 17/02/2005, Volume II5, No. 5943 Page 17

When Cortes invaded Mexico in 1519 he burned his own ships - cutting off any possibility of retreat. Ulysses had himself lashed to the mast to resist the lure of siren voices. And chancellor Gordon Brown did something similar in ceding control of interest rates to the Bank of England.

The point is this: tying one's own hands can be an act of strength. It can prevent you from succumbing to short-term pressures and ill thought-through compromises.

There are many ways to establish 'irrevocable commitments', as game theorists put it. Manifesto pledges are one example: they entice swing voters and commit governments to deliver. Pre-announcing NHS spending to 2008 is another example: making it politically impossible to cut the NHS budget in the mean time. But one of the most effective mechanisms is selectively to 'off-shore' functions altogether.

Think back to the mid-1990s.

Scarcely a week passed without one think tank or another banging on about the need for a national debate on NHS rationing. Then came the National Institute for Clinical Excellence, which - because it was at arm's-length from the Department of Health - injected transparency and killed the rationing debate dead.

By contrast, a few years before the inquiry on paediatric heart surgery at Bristol Royal Infirmary, a former permanent secretary told me that hospital inspections would be a terrible mistake.

'We would be inspecting ourselves' - and would be to blame for what we found. But by passing oversight of healthcare standards to an independent Healthcare Commission, problems are less likely to be brushed under the carpet.

That brings us to the foundation trust regulator, Monitor. By surrendering their legal power of direction over foundation hospitals, health ministers reduce their ability to succumb to temptation and may also over time reduce demands on them to do so. That seems to be the experience elsewhere in Europe. Of course the rigour and transparency of such a regime will often feel uncomfortable, as Bradford's travails testify. But let's not forget the dynamic that it seeks to replace.

I have seen it from both ends of the telescope. First was as a general management training scheme trainee in Consett, County Durham (just back from South America and my own 1980s version of The Motorcycle Diaries). Picking up the phone one morning, I found myself being grilled by health secretary Ken Clarke's private secretary about a press story that we were about to regrade our medical secretaries.

Wind the clock forward and I am sitting in the health secretary's office as Alan Milburn is instructed by the Speaker to appear in the Commons and take personal responsibility for malfunctioning hinges at Bedford Hospital's mortuary. And likewise a year later at Number 10 when prime minister Tony Blair is blamed for Rose Addis's treatment at north London's Whittington Hospital.

These are not accidental features of a system in which hospital accountability is hardwired to our health minister; they are the inevitable and dysfunctional consequences of it.

Hence the importance of foundation trust status and Monitor as partial circuit-breakers.

But does this go far enough? On balance, I think not. Three other functions now need putting on a transparent and permanent footing, insulated from the inevitable tendency continually to fiddle around in search of lowest common denominator solutions.

One is the regulation of anti-competitive provider behaviour, which might become the responsibility of the competition authorities as discussed in my last HSJ column (Opinion, 13 January, page 17). Another is patients' choice entitlements, which might be put onto a legislative footing. Third, crucially, is payment by results.

By abandoning the price=cost rule, the tariff system will give hospitals genuine efficiency incentives. But it also eliminates their inflationary pricing power, created by oligopoly plus information asymmetry relative to primary care trusts. It means PCTs can shift their attention from unit prices to volume, mix and quality of healthcare. And it underpins patient choice, in part because PCTs can be financially indifferent as to where their patients opt to be treated.

That is the theory. The practice depends on getting a lot of detail right: not paying for emergency readmissions; the right level of clinical specificity; standardised tariff-splitting; normative pricing for desirable additional activity;

periodic but not annual recalibration and rebasing, and so on.

Yet the understandable decision to slow implementation of the tariff for emergencies means that the elective tariff will be easily gamed.

PCTs will find that hospitals exposed to competition in their elective services will simply recover more overhead in their locally negotiated emergency prices. Half a tariff system doesn't make a coherent hospital payment mechanism.

So the time has come to consider vesting NHS tariff construction in an arm's-length technical agency. Its functions would be advisory, and it would not be an economic regulator in that the DoH would publicly set the financial parameters the tariff authority would work within.

However, like NICE, the assumption is that DoH would follow its recommendations unless ministers gave formal reasons for not doing so.

The agency might have a supervisory board comprised exclusively of NHS purchasers, or a balanced group of purchasers and providers. There are several international models to draw on: the Netherlands and the US's Medicare Payments Advisory Commission are two.

Given the weight that is to be placed on the tariff mechanism, we need more precision, predictability and permanence in its operation.

That is the bottom line.

For the next phase of NHS reform to succeed, we need a bit more lashing to the mast, and a few more burnt boats. Ulysses and Cortes point the way.

Simon Stevens is president of UnitedHealth Europe. From 19972004 he was the health policy adviser at Number 10 and the Department of Health. Simon_L_Stevens@uhc. com