'The NHS can't avoid political controversy, because what it does is too important, complex, and subject to debate'

It will be interesting to see how all the parties envision healthcare with Gordon Brown at the helm and the Conservative health 'white paper' on the table.

The difficulty all parties face is that there are few policy ideas to choose between. There are only so many ways you can steer the NHS, which has the scary potential to lose the election and sink the economy.

Whichever side you are on, you can't much improve the way the NHS is financed from direct taxation. The NHS enjoys the cheapest and fairest finance system in the world and no-one in their right mind would change it. The market failure inherent in competitive health insurance, which blights the US, is something for which we should all give daily thanks we don't have in the NHS.

Despite enthusiasm for a non-partisan board, it's hard to envisage big changes to how we govern the NHS. Whether it likes it or not, the Commons has to care what the NHS does, because it employs more than a million people; makes headlines because it cures and kills people; and has the power to make or break the national budget.

Even if governance is handed over to a non-elected body, that body will soon be as political as the Commons, if perhaps not strictly along party political lines. In other words, the NHS can't avoid political controversy, because what it does is too important, complex, and subject to debate.

You can't do much about the ageing population, unless you allow more young workers to immigrate to Britain. Migration from the EU may help a bit.

You can't keep wages down for any length of time. You might even have to put them up if you want to keep our (and other countries') nursing staff.

You might be able to change the name and rhetoric around service goals and targets (a welcome change for doctors at least). But you can't withdraw too many without risking plummeting and inconsistent standards, which will cause electoral bother. It is hard to see how the public will be happy, for example, if you start removing the two-week maximum for cancer diagnosis and the four-hour maximum stay in accident and emergency.

It is difficult to introduce performance-related pay without increasing management costs and alienating the unions. If you want to promise voters more frontline workers and fewer managers, this isn't the way to do it.

But if you want true competition in primary care, it might be worth examining Australia's model, where patients can pick and choose between any GP at any time, carrying a 'Medicare' card which entitles them to care.

It's great for patients, because they can visit one practice when they're at work or on holiday, another when they're at home, and choose specialised practices for different kinds of problems.

People who care most about continuity of care can stay with one GP - it's their choice, not the doctor's. It puts real pressure on GPs to stay on top of their game. Such a reform, however, would require a different kind of billing system (fee-for-service v capitation) and risks alienating doctors. So not for the faint-hearted.

You might be able to promote healthy competition between hospital providers but, as with railways, past experiments have not worked brilliantly, because there are too few hospitals to give consumers much of a choice. It's just possible that the private providers of elective procedures are having a positive effect on NHS providers with improved outcomes overall, though by no means certain. You could push this further; you'll need the courage to stop if it's proving destructive.

You should be able to keep improving procurement efficiency. One of the NHS's biggest strengths is its purchasing power. Walmart wouldn't hesitate to exploit this to the maximum and there's no reason why the NHS shouldn't do so either, so long as it doesn't compromise quality. This can extend to medical technologies, to dampen the year-on-year price rises above normal inflation. The same caveat on quality applies, of course.

A reform that would make the NHS easier to run is to divide the whole thing into one insurer-gatekeeper and many provider units. In most countries, this division of labour is normal: a national insurance body makes decisions about what mostly will be covered, while local providers deliver care with reasonable degrees of flexibility.

The NHS is unusual in that insurer and provider functions are still carried out at all levels, so frontline workers sometimes make on-the-spot coverage decisions.

This is like your mechanic having to mend your car and deciding whether your insurance should pay for it. Such an old-fashioned arrangement, a legacy from the NHS's pre-war origins, is inefficient and inequitable, resulting in postcode lotteries (the Herceptin debacle, for example). Sorting it out would make the whole system easier to manage.

There are many exciting IT innovations to be introduced which would please all punters: electronic patient records, seamless prescription and procurement systems, booking systems, to name a few big ones still waiting to happen.

And then there are the crowd-pleasers like cheap bedside internet, phone and music. Doing this across an organisation the size of the NHS is not easy to achieve and can cost the earth (£12.4bn so far), so you need to know what you're doing.

None of the options are partisan because in healthcare, economic realities swamp most ideological preferences. And the NHS is already suffering reform fatigue. So it will be interesting to see what the parties come up with to distinguish themselves. I know what I'll be looking for: not too many changes; a steady hand at the tiller, and a good head for economics to steer the NHS through increasingly difficult waters. Call me boring.

Dr Anna Donald is chief executive of healthcare information provider Bazian.