Leading article: Time for a fresh look at cancer screening

Should the NHS be screening for breast cancer? That may seem a question with a self-evident answer but the pages of the British Medical Journal have been crackling for the past six months with a dispute over the issue, since a Danish study questioned the benefits of screening. It emphasised the harm caused by over-diagnosis – where a cancer is detected that would never have become harmful – and the psychological stress, painful investigation and unnecessary treatment that follows. Supporters of breast cancer screening have howled in alarm at the damage such reports do to the uptake of mammography. Critics of the programme have accused them of substituting medical propaganda for proper scientific judgement.

In an attempt to resolve the problem, the BMJ commissioned Klim McPherson, a respected Professor of Public Health Epidemiology at the University of Oxford, to review the evidence. Alarmingly he has concluded that screening for breast cancer has only limited benefit, may harm individual women and is of "marginal cost effectiveness" for the community. The time has come, he suggests, for a serious scientific rethink of the benefits of the £75m-a-year NHS screening programme.

The Government should take note of this, particularly at a time when the nation cannot afford public spending which is ineffective. The danger is that it will not dare do so, for fear of the public outcry which would ensue – as happened in the United States when the US Preventive Services Task Force recently recommended that screening should not begin until the age of 50 (it was 40) and should take place only every two years rather than annually. The proposal provoked outrage from screening enthusiasts.

Consider the facts. There are both benefits and risks in screening the whole population for a particular disease when they reach a designated age. Common sense says that any kind of screening must be helpful, since it gives an advance warning and enables early treatment. But common sense is not always the most reliable guide to matters of science. The benefits easily outweigh the risks with some diseases, like bowel cancer, where diagnosis is straightforward and treatment is unambiguous and overwhelmingly successful. But with diseases like prostate cancer the diagnostics are highly unreliable and have, in the US, produced an epidemic of unnecessary treatment in a disease which can be so benign that many sufferers live for decades and die of something else before the cancer becomes anywhere near lethal.

With breast cancer, risks and benefits are more finely balanced. Detecting invasive tumours before they are clinically apparent is obviously a good idea. But screening will also detect tumours that are not going to cause trouble. Mammography itself is minimally carcinogenic. Intensive screening programmes have uncovered a new type of breast cancer – ductal carcinoma in situ – about which comparatively little is known, so that judgements on risks versus benefits are being made on statistics which are very imprecise. And screening all women aged 50-70 every three years is expensive; the money might be better spent on improved treatment.

Most alarming is Prof McPherson's suggestion that the case for an important national screening programme has been taken as an article of faith and not subjected to proper scientific scrutiny. Rather it has been led by "agenda-driven analyses" and "obfuscation from organisers of mammography services".

These are grave charges. The Government should commission a full and dispassionate examination of individual patient data from all recent studies to address the many unanswered questions. In the meantime, the NHS screening programme needs to be clearer about these uncertainties when communicating with people who need complete honesty if they are to make informed decisions about the benefits of screening.