For most types of cancer, surgery to remove a tumour is by far the most likely way to cure the disease compared to any other type of treatment. Surgery can also play a key role in relieving symptoms, even when the chances of long term survival are low.
Until recently, there has been little reliable information about the number of cancer patients in England having surgery. This is a problem – understanding why some patients get surgery and others don’t could help to improve standards of care and increase cancer survival in this country.
To address this deficit, my team at the NCIN have published a new report this month, which shows variation in the proportion of patients having major surgery around the country for a range of common cancers. In fact, our report looked at patients diagnosed with one of thirteen different types of cancer in England between 2004 and 2006.
Our report shows that surgical rates fall off steeply with the age of patient treated. And while this isn’t intrinsically surprising, the drop in some cancers begins in age groups as young as the 50s. This is a cause for concern. And, crucially, it raises questions about the underlying reasons for these variations and, in turn, what can be done to reduce them.
Inevitably, there may be reasons for variations by age and geography that are not directly related to the quality of care in our hospitals. Such reasons may include:
the stage of the disease at diagnosis,
late presentation by patients with symptoms,
patients – especially older ones – choosing not to have surgery,
different numbers of patients with other illnesses which mean surgery would be ill-advised.
But because complete data on these factors is not available, the report can’t attribute any of the variation to them, highlighting the importance of NCIN’s ongoing work in improving the information collected about cancer patients.
Despite all these caveats, the new facts and figures suggest that some parts of the country may have different rates compared to other regions. This report provides the basis for further analyses to try and understand what the underlying causes of these differences are.
This is a good basis for cancer networks to examine why they might have lower surgery rates than their neighbours. So publishing this type of analysis could actually help to make things better.
My analyst colleagues at the NCIN linked routine hospital data known as Hospital Episode Statistics (HES) with cancer registry data, to create a snapshot of the number of patients having major surgery as either an attempt to cure cancer or as palliative care to relieve symptoms.
All this information is collected by hospital trusts and cancer registries. The process relies on each trust accurately recording and assigning a code to the type of surgery carried out so that they can claim reimbursement for every operation. The recording of data on surgical operations for patients is one area where clinicians feel improvements in data collection is needed.
It is, however, the only available national data on surgical operations and procedures in England and analyses such as this highlight the importance of ensuring data is recorded accurately throughout NHS trusts in order to gain a better understanding of treatment for patients.
This new report is the first time that so much information on many common cancers has been collected, and it does provide an insight into cancer surgery in England.
There is wide variation in the proportion of patients having surgery for different cancers. For example, nearly 85 per cent of breast cancer patients have surgery, compared with fewer than 10 per cent of lung cancer patients.
We can’t make an international comparison with these figures, because there’s no reliable equivalent data covering most cancers in other countries. But we do know the surgical rate for lung cancer in England is definitely lower than published in a number of studies carried out elsewhere. Many other countries report lung cancer surgery rates of over 20 per cent – double the level that we found in our study.
The report also found variations in rates of surgery for most cancers according to where patients lived (the exceptions being breast, uterus and liver cancer).
In lung cancer, for example, patients living in Merseyside and Cheshire, Birmingham, Greater Midlands and East Midlands had rates of surgery significantly higher than the average. In contrast, patients living in Lancashire and South Cumbria, North Trent, Mount Vernon, Sussex, Kent and Medway and Essex had significantly lower rates of surgery.
The variations could not be explained by age differences in these populations but due to the lack of complete data for other factors, including stage of disease at diagnosis, we have not been able to show what the underlying causes of this variation are. Improvements in the availability of cancer staging data nationally are being made and this will allow future analyses to investigate how differences in stage of disease impact on this variation.
As I mentioned above, our report also shows that surgical rates fall off steeply when patients are older. This may not seem particularly surprising, but for some cancers the rates begin to fall in patients as young as 50. There was a particularly steep decline in middle-age for ovarian, kidney and cervical cancer. But there could be several reasons for this, including differences in the types and stages of cancers presenting in the different age groups.
For example, in cervical cancer, younger women’s cancers are more likely to be picked up through screening and therefore may be detected at an early stage. Early stage cancers are suitable for surgery whilst more advanced disease is treated with chemotherapy and radiotherapy. Younger women also tend to opt for surgery as radiotherapy can cause long term scarring.
The report also looked at whether social and economic circumstances might influence whether cancer patients have surgery. Perhaps surprisingly it found there was no major fall off in the rates among patients living in poorer areas. The exception was cervical cancer, where rates were 10 per cent lower among the most deprived patients compared with the most affluent, however, this may reflect differences in stage of disease at presentation.
To understand these variations we need to identify any other factors linked with age that stop patients having surgery. In some cases surgeons may not want to operate on older patients, placing a greater weight on the potential harms of surgery than on its benefits. Or patients themselves may be choosing not to have surgery.
Our report is only the start of an ongoing programme of work to uncover any variations in use of surgery and other cancer treatments.
For a start, high-quality information needs to be gathered on the stage at which a patient’s cancer is diagnosed and any other illnesses they have that might affect their suitability for surgery. This would allow for a fairer ‘like-for-like’ comparison between hospitals. We also need to explore whether some of the variation in the rates of surgery can be explained by variations in the use of alternative treatments (especially radiotherapy).
Another issue is that there is no international agreement on what the ‘ideal’ surgery rate is for most cancers. Nor do we know enough about how overall rates of surgery in populations of patients with different cancers impact on the survival rates, or on the risks of serious side effects or even death as a result of the treatment.
Examining the links between treatment rates and outcomes has already begun and will form a major element in the joint work the NCIN is doing with the Department of Health and Cancer Research UK on international benchmarking.