The most commonly cited source of international cancer survival rates is the EUROCARE series of studies, one of the largest analyses of European cancer data, the most recent of which (EUROCARE-4) covered patients diagnosed in 1995-99 and 2000-02. It showed that survival for the four most common cancers (lung, breast, prostate and bowel) and for ovarian cancer was lower in the UK than in much of the rest of Europe, and that the UK had lower overall survival rates compared to countries with similar expenditure on health.
However, there are two problems with using EUROCARE data to judge the government's performance on cancer.
The first is the time lag associated with studying five-year survival rates – patients in the study were all diagnosed either before or in the very early days after publication of Labour's first cancer strategy, the NHS Cancer Plan, in 2000.
The second is that few European countries have the fully comprehensive cancer registry data that we have in England. Coverage of cancer data in Germany, for example, amounts to only 1 per cent of the population, and France only approximately 10-15 per cent. To overcome this, more detailed work on international comparisons in cancer is being developed by the government in partnership with six other countries.
Despite these limitations, EUROCARE does paint a broad picture that survival rates, although improving, have been relatively poor compared to other countries. So why is this? And what is needed to improve cancer outcomes?
One of the major areas that we now know needs greater attention is early diagnosis, at least for some cancers, as shown in a series of papers published recently in the British Journal of Cancer. One way to show this is to look at survival rates after only one year, as people who die within a year were probably diagnosed at a very advanced stage.
For example, one study compared survival rates for all cancers in the UK with those in the Nordic countries (which have some of the highest survival rates in the world and have comprehensive data), and showed that our one-year survival rates are 10.8 percentage points lower. Excluding those people who die within a year, as a way of attempting to compare patients who were not diagnosed late, our five-year rates were only 3.6 percentage points lower, suggesting that late diagnosis could be to blame for much of the overall difference.
Although early diagnosis was mentioned in the Cancer Plan, the major focus in 2000 was on cancer treatment, particularly speeding up access to treatment once patients had been diagnosed. There have undoubtedly been very significant and important improvements in cancer services: cancer screening has been extended and improved, patients are being treated faster, more patients are being treated by specialist surgeons, and treatment is better co-ordinated through multidisciplinary teams.
Yet it's only relatively recently that early diagnosis of symptomatic cancer has become a top priority, following the publication of the Labour government's Cancer Reform Strategy and the establishment of the National Awareness and Early Diagnosis Initiative (NAEDI). It's too early to tell its impact, but the EUROCARE data and the work of NAEDI suggest some clear lessons for the next government.
First, it should exercise caution when using inevitably old data, such as five-year survival rates, to claim new achievements or blame recent failure. Second, it should continue to support more detailed studies both of UK data and of UK performance in comparison with other countries to understand more clearly what is driving cancer outcomes. And third, initiatives to improve early diagnosis need to be at the top of the Secretary of State's to-do list.