In 2007 the UK government published its landmark Cancer Reform Strategy. As with previous government strategies, such as the 2000 NHS Cancer Plan, the report aimed to tackle the fact that the country’s cancer survival lagged behind other similar countries.
But unlike previous cancer strategies, which focused on improving patients’ access to the latest cancer treatments, the new report highlighted the fact that the UK’s underperformance was also partly because cancers are generally diagnosed later in this country than in many other countries.
This was a significant shift in emphasis. In fact, experts since estimated that up to 10,000 deaths could be avoided every year in England if we raised our game to match the best in the world.
But what’s the cause of this discrepancy? Do the differences still exist? And what practical steps can be taken to bring the UK in line with our neighbours?
In response to the strategy, a group of world-class international scientists, including several of Cancer Research UK’s key researchers, set out to answer these questions by forming a collaborative group called the International Cancer Benchmarking Partnership, or ICBP.
The ICBP is split into five separate modules (of which more later). The first of these aims to update and compare survival rates between different parts of the world.
Today the group working on module one has published research, in top medical journal The Lancet, looking at patients diagnosed between 1995 and 2007 in six countries around the world. The new figures give a concrete foundation for the ICBP’s future work, establishing the current scale of the problem. And they allow the effectiveness of any action taken over the next few years to be properly measured.
The analysis published today finds that survival rates for four different cancers – lung, bowel, breast and ovarian cancer – have continued to improve across the world, and that the UK has begun to close the gap in some, but not all of these diseases. Overall though, it confirms that the UK is still lagging behind other developed countries, and that there’s much to do to improve things for UK cancer patients in the future.
This finding is an important one, and it’s worth having a look in detail at what the researchers measured, and what they found.
What is ‘survival’ and how is it calculated?
Cancer survival figures give an indication of how many patients, on average, are still alive a specific time (usually one or five years) after being diagnosed.
Survival rates include people with cancer who die of any cause – for example, old age, treatment complications, accidents, etc. So to take account of the fact that some patients will die of other things, statisticians calculate a measure called ‘relative survival’. Because it takes other causes of death into account, relative survival is extremely useful for making international comparisons.
Survival statistics are related to, but not the same as, cancer ‘mortality’ rates – which measure ‘people who die of cancer’ (as opposed to ‘people with cancer who die’ – cancer survival).
Compared with survival rates, mortality rates are relatively easy to obtain, since the data is drawn from the information on patients’ death certificates. But mortality rates don’t tell us anything about how quickly after diagnosis a death occurs, so they don’t give us the whole picture.
Relative survival, on the other hand, allows statisticians to say something about the time between diagnosis and death, and therefore the influence of a country’s health system on how a patient fares.
But calculating relative survival rates is a complicated business, requiring high-quality data from cancer registries (organisations that track cases of cancer), as well as death certificate information. And you need to link a person’s record of being diagnosed to their death certificate.
What did the researchers calculate and why?
In the Lancet paper, the ICBP module one team, led by Cancer Research UK survival expert Professor Michel Coleman, calculated one-year, five-year, and “conditional-five-year” relative survival rates for four different cancer types across six different countries.
One-year survival is an important measure. For many cancers, lower than average one-year survival rates tend to suggest people are being diagnosed later, when treatment is often less effective.
Five-year survival tells us more about the effects of treatment, and is a widely used measure of cancer survival.
In many ways, the most interesting measurement they calculated was the “conditional-five-year” survival – the proportion of patients who survived the first year, who are still alive at five years. This gives a sense of how good a country’s quality of long-term cancer treatment and other care is, in a group of patients who should have better outcomes as they have managed to survive the first year.
The team looked at survival in four cancer types. They chose lung, bowel and breast cancer because together these cancers account for over four out of ten cancer deaths, but also because they tend to have differing prognoses (poor, intermediate and good, respectively). Ovarian cancer was also chosen as an example of a less common cancer that is often diagnosed late.
The researchers compared data from the UK (excluding Scotland), Australia, Canada, Sweden, Norway and Denmark.
What did they find?
The good news is that survival rates are improving in all of these countries, in all four cancer types – as you can see in the graph to the right (see figure 1).
Encouragingly, the team also found that the rate of improvement in the UK was, in general, slightly better than in other countries, particularly in breast cancer, where we have made extremely good progress.
However, taking bowel cancer as an example (figure 2), relative survival in the UK is the least impressive of all six countries covered in the research, despite the fact that it’s improved by nearly 6 percentage points from what it was in 1995-1999.
Across all countries bowel cancer was, in fact, the ‘most improved’ cancer studied. However, the UK though bucked this trend and the biggest improvement was in a different area – breast cancer. This is probably because the other countries already had good breast cancer survival figures.
The data also showed that, of the three UK countries included in the study (England, Wales and Northern Ireland), Northern Ireland showed the best survival rates in all four cancer sites studied (figure 3).
What does it all mean?
This is the very first paper from one of the ICBP, and it establishes an up-to-date picture of how cancer survival rates differ between countries. It provides an extremely solid platform on which future work can be based. And as the authors write, the results show that “data quality and changes in classification are not likely explanations” for what they observed.
But as such it doesn’t, on its own, tell us anything about why these differences are occurring. This is work that the other groups will hope to answer over the coming years.
The module one group plan to publish several more papers in the near future, including a more detailed analysis of the data they’ve collected so far. And the other ICBP groups are already examining a range of potential reasons for these differences in survival. These include:
- late presentation and more advanced stage at diagnosis;
- differences in treatment;
- awareness and beliefs about cancer in the general public, and amongst GPs; and
- the delay from first symptoms to being diagnosed and treated.
The work of these ICBP groups, which Cancer Research UK is supporting, will ultimately help politicians and healthcare providers improve the outlook for cancer patients of the future.
Data collection is vital to demonstrating that cancer survival is improving
But there’s one final, and crucial, thing that these results provide us with – and that is evidence. There has been much talk in recent years of the need to protect frontline services in the current financial climate. But in order to carry on making progress as the country has done, we also need to ensure that the collection and analysis of information about cancer is also protected.
As our director of health information, Sara Hiom, told the media this morning, “reliable data – which are consistent across the country – are crucial to understanding the extent of the problem, and identifying the causes of survival gap within the UK and compared to other countries”.
As the government gears up to publish its new ‘refreshed’ cancer strategy in early 2011, it’s a point we urge them to keep in mind.
Coleman, M. et al (2010). Cancer survival in Australia, Canada, Denmark, Norway, Sweden, and the UK, 1995–2007 (the International Cancer Benchmarking Partnership): an analysis of population-based cancer registry data The Lancet DOI: 10.1016/S0140-6736(10)62231-3