You may already have seen the moving story of NHS surgeon Gordon Matthews and his wife Sue on the BBC News website. The right to top up treatment is a genuinely hard subject, even before you start giving it a human face. When policy meets reality and personalities, the clash can be deafening.

When you know you are facing death, the feeling must be desperate.

This is no more and no less true for Gordon and Sue Matthews than it is for any of the other hundreds of people who have faced two stark choices in a row: whether to spend thousands of pounds of their own money to top up their NHS treatment with expensive anti-cancer drugs that may extend life my a few months; and then whether to fight the system which may be attempting to stop them doing it without losing their NHS free care.

Why is the system trying to stop them?

In essence, because it is trying to enforce equity of provision. We all pay different amounts into the NHS, but the basic offer of this system is that we should all be receiving the same treatment.

The idea behind this unpopular policy is that if you have a rich patient next to a poor patient on a ward, the former's ability to buy expensive drugs which the latter cannot and the NHS does not fund creates a fundamental health inequality (assuming of course that the drugs work).

So there is a reason behind the policy. You may disagree with the reason. But there is one, and it is principled, and opponents of the current enforcement of policy should at least engage with the argument.

In ethical terms, perhaps the question we should ask ourselves is 'does the end justify the means'? For Gordon and Sue Matthews and hunderds like them, the end of topping up is an extension of life and a temporary staving off of death. The means is using one's personal wealth. For NHS policy as it currently stands, the end is to reduce inequality. The means is draconian and appears uncaring. And we know that ends should not justify means.

The by-product is less equality. Yet our society is wildly unequal already.

Bentleys are not equitably distributed through society; nor is Petrus; nor is access to good-quality education; nor is access to reasonable and affordable housing.

The former two could prove determinants of ill-health status, but the latter two are key determinants of good health and life chances.

Here we get into the territory of translating policy into practice. Would it be more unacceptable to have the policy, but for NHS trusts to turn a blind eye or never enforce it?

Obviously, that would be officially unacceptable. Yet it is clear that there is wide variation from oncologist to oncologist, unit to unit and trust to trust on the enforcement of this policy. It's called localism.

We could either form a cancer drugs network, funded through a ring-fenced central budget to purchase at national scale, getting the pharmaceutical industry's agreement that the drugs will be reimbursed on a strict, independently monitored performance basis. This would be the statist option, and requires funding.

Or we could stay with the idea, harsh as it may seem, that we need to draw a price line beyond which the NHS does not reimburse treatments - and know that in doing this, there will be human distress caused by people not being able to access treatments from which marginal benefit might be gained at a cost the NHS will not pay. This would be the status quo option, and requires no change.

Or we can let people top up their care if they are wealthy, and extend our sympathies to those who are not wealthy. This seems to be the dominant suggestion about top-ups.

If we took option one, then what would we not fund? From where would we take the money? Prison health? Mental health? Falls prevention services?

Or do we raise taxes?

UPDATE 6.45 pm GMT - see www.healthpolicyinsight.com/?q=node/112