OSB archive
OSB archive

Cancer screening: the facts

Jonathan Wood

There have been a number of stories about cancer screening recently, not least with the reams of coverage of Jade Goody’s death and now her funeral.

Her tragic case seems to have increased the number of young women going for cervical cancer screening, but there has also been criticism of the information given to women undergoing breast cancer screening. What are we to make of all this? Thankfully, Jane Green of the Cancer Epidemiology Unit at the University of Oxford can act as our guide.

OxSciBlog: What differences have screening for breast cancer and cervical cancer made?
Jane Green:
Breast screening is thought to save 1,400 lives a year in the UK, a reduction in mortality of about 25% in screened women. UK cervical cancer death rates have fallen by half since screening was introduced in 1988, with an estimated three-quarters of cancers that would otherwise occur (as many as 4,500 a year) being prevented by screening.

OSB: How can research determine when screening is effective, improves outcomes, and is cost-effective?
JG: This kind of research can be difficult, very time-consuming and expensive. But it is necessary if a screening programme is to be based on good evidence and to do more good than harm.

We need to make sure the screening test used is the most reliable test. Then we need to try out the test in a pilot programme to make sure that it works in practice, that the programme is feasible from a staffing, infrastructure and financial point of view; and that it is effective in reducing later serious disease or death.

The research also has to take into account coincidental changes in the underlying rates of the disease, and changes in survival because of new treatments.

The best way to test a screening programme is to do a randomised trial and the results between screened and unscreened groups compared over many years. Financial costs of screening have to be compared with the cost of not screening (where more people may go on to develop serious illness later) and with other health care priorities. Finally, when the programme is set up it is crucial to evaluate its performance at every level. 

OSB: Are there problems with false positives and worrying people unnecessarily?
JG: Yes. This is an inevitable part of screening. In an ideal world we would have a perfect screening test – one which picked up every person with early disease which we could usefully treat, but no-one who did not have the disease.

In practice there will be both 'false positives', people who test positive, but who do not have the disease; and 'false negatives', people who do not test positive when screened but who do in fact have the disease.

False positive tests cause, at the very least, anxiety and the need to have further tests; and, at the worst, unnecessary and often unpleasant treatment, because screening will detect some early stage disease which would not go on to cause problems later.

For the individual with a positive screening test, there is therefore always a chance that they are one of the 'false positives'. Some people will find this out when they are called back for further tests and find they do not after all have cancer. Some people with early disease detected by screening will never know whether their cancer treatment was necessary. What they do know is that overall, screening and subsequent treatment saves lives and that theirs may be one of these.

OSB: Should younger women be screened for cervical cancer than currently?
JG: This is not an easy question to answer. Cervical cancer is extremely rare in very young women (there are about 50 cases a year in the UK in under-25 year olds, and about 5 deaths) and abnormal cell changes on the cervix are quite common, but often resolve themselves without treatment.

For women under 25, the experience from the English screening programme has been that on balance, the harm caused by detecting and treating cell abnormalities in many screened women (some of whom may have later problems with pre-term birth because of damage to the cervix from treatment) outweighs the benefit of picking up the very rare cases where treatment is necessary. This is why it was decided to raise the lower age limit for routine screening from 20 to 25 in 2003.

The difficulty is of course that this decision is based on the balance of risks and benefits for the population as a whole. For the individual it is different. Some young women will benefit from not being screened, and some young women would benefit from screening.

Cervical screening is designed for women with no symptoms. A woman of any age who is concerned she may have a problem can of course ask for a test through her GP, family planning clinic or a genitourinary medicine or sexual health clinic.

OSB: What should women do in the face of the mixed messages they hear?
JG: It is important to remember that the cancer screening programmes in the UK are based on good scientific evidence. There is no serious argument about the overall effectiveness in terms of lives saved.

The debate is about the difficult issue of balancing the benefits and harms of screening: and screening always has both. Some women will have cancer that is missed, and some will be diagnosed with cancer that might not have progressed if left untreated.

Assessing the numbers involved is extremely difficult and involves making assumptions: we cannot know for certain how many cancers would not have progressed, or how many cancers would have been detected without screening.  Estimates of the numbers of cancers 'over diagnosed' vary widely and some of the recent debate has been based on extreme values.