Arguments about what we get for our billions of NHS pounds rumble on. The Department of Health's latest analysis - noted here last month - claims the extra billions have essentially been spent on extra staff. We could take it to be a good thing - if the job of the NHS was simply to provide employment for doctors, nurses and managers. But wouldn't it be better to know what outcomes it bought other than extra labour?
Most of what the NHS does has never been subject to proper clinical and economic trials. But recent research by economists at York University - using data from the DoH's national programme budget project - has made important progress.
There have been numerous studies over the decades, and around the world, using a variety of statistical techniques to establish what factors explain changes in health over time or variations in health across countries or areas. Broadly, such studies confirm a positive link between healthcare spending and health. But these studies tend to be on a macro level, focusing on the broad determinants of health..
Any system that allocates more resources per capita to areas with poorer health inevitably throws up the simple (cross sectional) correlation that areas that spend more have poorer health. On the face of it, not a good result.
But the York researchers have managed to disentangle this problem, in part by standardising variations in health need and through more sophisticated econometric techniques.
So far they have looked at cancer and circulatory diseases. These accounted for around£10.4bn of primary care trust spending in 2005-06, with average per capita spending of around£83 for cancer and£124 for circulatory diseases.
Their findings suggest that a 10 per cent increase in cancer spending results in a near 5 per cent reduction in cancer deaths in the population. For circulatory diseases, a similar increase in expenditure is associated with a 14 per cent reduction in deaths. So, more money in appears to mean more health out.
Using years of life lost rather than standardised mortality ratios as their outcome measure, the researchers were also able to estimate how much it cost to generate a year of life. For cancer this turns out to be in the range£9,100 to£23,500, and for circulatory diseases,£6,600 to£10,200. These are cost per life years, not quality adjusted, but compare favourably with the National Institute for Health and Clinical Excellence's 'acceptance range' of between£20,000 to£30,000 per quality-adjusted life year.
John Appleby is chief economist at the King's Fund.