Earlier this month another health authority said it was contacting thousands of women for cervical smear tests after it became clear that its system for recalling women who had had borderline test results had broken down. In south London and Surrey, nearly 4,000 women must undergo the anxiety that the tests produce. Some will have treatment as a result, including women whose signs of pre-cancerous cell changes were slight. Sometimes the cells can revert to normal. These women may wonder how necessary that treatment is.

Increasingly, patients are challenging doctors' decisions. Last autumn a charity was launched to raise money for legal action against surgeons who remove a woman's womb or ovaries without consent, while an organisation called Rage has been set up to seek compensation for women who have suffered from the side-effects of radiotherapy for breast cancer.

The battle is not only over women's bodies. The death last summer of six- year-old Daniel Stoneman, from Devon, whose mother had opposed further cancer treatment for his muscle tumour, brought this emerging debate into focus.

How much cancer treatment is enough? And do doctors know when to stop?

Daniel had a rare cancer that had already paralysed part of his face, and doctors had planned 30 sessions of radiotherapy, which they believed would give him a 30 per cent chance of recovery. Legal action against his mother was considered because she refused the treatment. She argued that each session would have required a general anaesthetic and that the treatment risked damaging his brain. Angela Stoneman said she would rather her son die in peace at home than suffer any more in hospital.

The question of when to stop treatment is one faced by many people with cancer, particularly if the disease is advanced. The subject is no longer taboo; a US survey found that in 1961 only 12 per cent of doctors told their patients if they had cancer; by 1979, 98 per cent did so. But while medical developments have dramatically improved the chances of survival for those with some less common cancers, such as childhood leukaemia, and improved the outlook for others when cancer is diagnosed early, beyond this doctors can claim only modest success: most advanced common cancers, such as those of the breast and lung, are incurable.

People with advanced cancer can nevertheless be subjected to treatments such as chemotherapy and radiotherapy, which may have severe side-effects and only rarely prolong life. Now pressure is growing for greater scrutiny of this palliative care, where the aim is to alleviate symptoms rather than provide a cure. 'Quality of life', a phrase used by Daniel Stoneman's mother, has become a catch-phrase among concerned researchers. Although this may seem a relatively humble goal for health professionals fighting cancer, it recognises the importance of normality and acknowledges that more is not always best.

'There is a danger that in some areas people will be over-treated, in some cases for the kindest of reasons,' says Dr Chris Williams, consultant oncologist at the Cancer Research Campaign's Wessex Regional Medical Oncology Unit at Southampton General Hospital. To provide an example of over-enthusiastic treatment he cites Finnish research which indicated that women with advanced breast cancer were being treated with more intensive chemotherapy the iller they became. Women who died during the five-year study period had more treatment than survivors, and 30 per cent of the women who died had had very recent chemotherapy.

Pressure to treat comes from multiple directions. Patients and their relatives may be desperate for a last resort; doctors may be keen to enrol patients into clinical trials aimed at finding newer and better treatments; patients are aware of the pharmaceutical advances regularly publicised by the drugs industry.

There is also an element of 'hope over expectation'. Although doctors no longer withhold a diagnosis of cancer, they may in effect 'censor' information, giving patients the impression that the situation is less serious - or more hopeful - than it is.

'If you haven't said 'it is incurable, we cannot stop you dying', patients may get the impression they may still survive,' says Dr Williams. 'If the doctor starts talking about a 5 to 10 per cent 'response rate', the patient may still think there is some chance.' In fact, the response rate refers to the percentage of cancers that shrink by half or more during treatment. People who 'respond' are likely to report improved quality of life, but this is remission, not cure.

Dr Williams believes it is important for doctors to take the time to be honest with their patients, even if it means painting a bleak picture. 'People are saddened by it, but often not surprised. Very often they will say, 'Thank you for telling me, at least I know where I stand now.' '

Not all cancer care is 'over-treatment'. Dr Mike Richards, clinical director of oncology at Guy's and St Thomas's Hospital Trust, London, says: 'On an individual level there are almost certainly cases where a compassionate explanation of a person's circumstances would be better than more chemotherapy. But on a national level there may well be people who would benefit from chemotherapy to whom it's not being offered.'

In a study carried out by Professor Robert Rubens at Guy's and St Thomas's, researchers looked at successive treatments of chemotherapy for more than 700 women with advanced breast cancer. They found 34 per cent of the women responded to a first course of chemotherapy, for an average of eight months. The outcome after second- and third-line chemotherapy was significantly worse; 16 per cent responded, for an average of just over two months. The study could not predict which women had the greatest chance of benefiting from treatment.

Dr Richards says the findings have made him more cautious about offering multiple chemotherapy, although some of his patients still benefit from successive treatments.

Dr Maurice Slevin, consultant medical oncologist at St Bartholomew's Hospital in London, believes that cancer is under-treated rather than over- treated. The advent of new drugs, he says, means that chemotherapy can be given with very few side-effects.

People forget that cancer itself has side-effects, he points out. 'Cancer makes them feel sick and can cause severe pain, shortness of breath and loss of weight. If you give chemotherapy and the side-effects are less than the symptoms, overall the patient benefits.' He argues that the newer drugs used in cancer therapy do not have the side-effects of hair loss, nausea and vomiting and immune suppression common with older drugs. If modern anti-sickness drugs are also used, nausea and vomiting can virtually be eliminated. However these drugs are expensive, and in many instances doctors cannot use them as much as they would like.

Dr Slevin and his colleagues carried out a study, funded by the Imperial Cancer Research Fund, which showed that the thinking of people with cancer will often differ greatly from that of their doctors. Patients quizzed about hypothetical treatments with chemotherapy were much more likely to opt for radical and intensive treatment with minimal chance of benefit than were their professional carers.

To complicate matters further, some studies that have tried to measure the quality of life in people with advanced cancer have come up with unexpected results. A Scandinavian study of people with advanced colorectal cancer found that those who received more intensive chemotherapy reported better 'quality of life' despite having more severe side-effects.

Oncologists recognise that more research is needed. At Guy's and St Thomas's, a team headed by Professor Rubens is halfway through a major audit of appropriate use of palliative treatment in the South-East Thames Regional Health Authority. The project is looking at several common cancers. 'We have to find out if what's been going on all these years is worthwhile in terms of benefit to the patient,' says the project co-ordinator, Kieren Towlson.

The author is a doctor.