There is something very seductive about having a “check-up”. It seems to offer reassurance in exchange for anxiety. So why is it that more and more doctors are questioning the use of screening tests to detect problems in otherwise healthy people?

This may sound counter-intuative. But writing in the October issue of the American Journal of Urology, a group of American medical researchers dismissed the blood test used as a standard screening test for prostrate cancer as all but a waste of time. Not only was it useless, they argued, but it could even be harmful, triggering a string of unneccesary operations, concluded the researchers from Stanford University in California.

It is not only the PSA (prostate-specific antigen) test that is under scrutiny. At my GP surgery in Glasgow , I have lost count of the number of times a perfectly well person has asked me for a check up. But like many others in the medical profession, I believe that certain check-ups can do more harm than good.

Let me be clear: I am not talking about people with symptoms, who are rightly concerned and want information or a diagnosis. My doubts are over those people who have no symptoms of any illness- just a desire for a ’doctor’s check up’ and a rubber stamped signature of wellness.

Isn’t it better to find problems early? Surely it makes sense to search pro-actively for them? And isn’t that the way medicine is headed. We have ’wellness’ clinics, pay-on-demand CT body scans that claim to be the ‘ultimate health check’ by detecting cancer of many different types early, and a burgeoning private healthcare screening industry. As an advocate of health, shouldn’t I be pleased that people want to spend their money in this way?

The answer is not straightforward. Far from being a hodge-podge of tests done randomly, looking for disease in a person who has no symptoms of having any should be a rigorous scientific discipline.

In 1968, the World Health Organisation published Principles and Practice of Screening for Disease, by James Wilson, then Principal Medical Officer for the Ministry for Health in London, and Gunnar Jungner, Chief of the Chemistry Department in Sahlgren's Hospital, Sweden, and is fully available on it's website (www.who.it/en/) In short, they said that there is no point - medically or economically- in searching for a disease that you cannot do anything about. Before you start screening for a disease, you should know and understand the natural course of it. There should be a period of time where the disease is present yet it does not produce any symptoms, and where finding it at this early stage will bring more benefit to the person than it would if you found the disease at a later stage. The screening test should then be repeated at intervals, in accordance with what is known about the natural history of the disease. There should be adequate resources for doing the tests and for dealing with the results. And importantly, the person should gain overall: the process of finding and treating the disease should not be worse than the adversity of the disease itself.

These criteria may seem clear and sensible, but they describe the perfect screening test- and frankly, the perfect screening test does not exist. If it did, then screening tests would pick up major illness before they became serious, and it would be within our power to reverse or prevent disease with a harmless check-up.

But the truth is that all screening tests are fallible, and all, therefore, have the potential to cause harm. For example, a screening test may pick up a serious disease early. But it could also detect a minor abnormality, of far from certain significance. Such small abnormalities are common, but, once found, are difficult or impossible to ignore. So a problem that was not serious and never going to become so is picked up, leading to a cycle of testing, examination and anxiety. A well person becomes a patient.

Even teaching regular breast and testicular self-examination are no longer recommended. A large, well designed study published in the Journal of the National Cancer Institute in 2002 compared women in China who were taught breast self-examination with those who did not. The researchers found that there were no differences between the groups in the amount of death from breast cancer, but there were more biopsies and anxiety in the self-examination group. It seems that awareness of normality together with prompt medical advice if an abnormality is detected is just as effective, and less potentially stressful, as teaching women to self examine their breasts.

The truth is that few blood or urine tests can detect disease in a well person that would be amenable to a successful pre-emptive strike. Those that are – for example, checking blood pressure, measuring cholesterol levels or glucose, and screening newborn babies for certain rare metabolic diseases are already available on the NHS.

The majority of tests carried out in the private sector as ’health check-ups’ - have little credible evidence supporting their ability to find definite problems. Yet according to ‘Which’ magazine earlier this year, the UK spends £65 million annually in the private sector on check up tests. A cynical onlooker might think that the only beneficiaries are the ones who are being paid to do the tests.

Prostate cancer screening using the Prostatic Specific Antigen (PSA) test is even more contentious. PSA is measured using a blood test, and is extremely useful when used to monitor the treatment of prostate cancer, the second most common cause of cancer death in men after lung cancer in the UK.

However, around two thirds of men with high PSA tests will not have prostate cancer (a PSA can be raised for many other reasons), and some men with prostate cancer will have a normal PSA. The test, as it currently stands, is not capable of distinguishing the aggressive cancers from the slowly growing tumours which men die with, rather than from. Yet even a slightly raised raised result may lead to anxiety, and further tests including a biopsy.

Professor Malcolm Law, of the Wolfson Institute of Preventive Medicine is a critic of using PSA tests for screening and outlines the dangers. “Some men, diagnosed at screening, will receive treatment that is unnecessary, because the cancer is too extensive to be curable, or because the cancer would have been indolent and insignificant anyway” he says. “Up to two thirds of men getting radical surgery may become impotent or incontinent as a result of the surgery. And there are no grounds for assuming that there is any benefit in terms of cancer survival to that patient. ”

While over 70 % of men over 80 will have evidence of prostate cancer, only 3% will die as a direct result of it. The reality is that there is no good evidence that the treatment of screen- detected prostate cancer reduces deaths from it.

In the UK, there is no formal prostate cancer screening programme, but men are entitled to request a PSA test on the NHS. The NHS is also funding a £14M trial called 'ProtecT' (Prostate testing for cancer and Treatment). It should reveal more about whether or not treating screen-detected cases using PSA testing actually saves lives by randomly assigning – the ‘gold standard of clinical trials’ - the three treatment options of surgery, radiotherapy or active monitoring to patients. Crucially, the trial, run by Professor Hamdy in Sheffield, Professor Jenny Donovan in Bristol and Professor David Neal in Cambridge, is also assessing the potential harm that the testing programme can bring. However, both this, and a European study (ERSPC - European Randomized Study of Screening for Prostate Cancer) asking similar questions about the effectiveness of screening, are not due to report for several years.

In the meantime, advocates of PSA screening say, to paraphrase, that men should be ’better safe than sorry’. They believe that men should be encouraged to have the test as part of a bigger push to persuade men to take responsibility for their health.

Roger Kirby, Professor of Urology at the University of London is the author of Men’s Health (Isis Medical Media, Oxford) and Chairman of the Prostate Research Campaign (PRC-UK), and has a specialist interest in the prostate. “We believe in a pro-active approach to men’s health’ he says. “We accept that PSA testing is not a perfect test, but it is a good indicator that something may be amiss in the prostate. We use the PSA test to decide whether or not a biopsy of the prostate is indicated - and again accept that the biopsies themselves are not 100% accurate. But in our experience, many men who proactively seek evaluation of their prostate have many other concomitant conditions such as high blood pressure, diabetes and abnormalities in blood lipids. At the same time as offering PSA testing, we can provide other important men’s health messages about refraining from smoking, sensible drinking and exercise.”

What would he say to a man considering a PSA test? “We say that the trials to confirm that PSA screening saves lives are in hand. Currently, 10,000 men in the British Isles each year die of prostate cancer. If we wait 5 years for the results from these studies to be made available, then 50,000 men will probably go prematurely to their grave. Every case of prostate cancer goes through a curable phase before becoming incurable, and PSA testing can help to detect these tumours early. Not every raised PSA needs surgery – active monitoring can be a good option - but knowledge of it is power to decide on therapy or surveillance. So I myself have an annual PSA test and recommend that others do so.”

But I believe there are dangers here. Even with the best of intentions, this is a test with deep and long lasting repercussions, and, at the moment, the ability of PSA testing to provide a benefit for that person is far from proven. Men have a right to know about the divisive uncertainties of PSA testing before they consider having the test.

The problem for doctors is that they can be tempted to protect themselves from malpractice suits by carrying out tests rather than informing patients, without bias, of the risks and benefits of a test.

PSA is not the only cancer screening technique to have critics. Even well-established screening programmes like the Breast Screening Programme are not universally acclaimed. Running since 1990, their website (www.cancerscreening.nhs.uk) says that they have screened 14 million women and detected more than 80,000 cancers. The number of breast cancers being diagnosed as a consequence of breast screening has undoubtedly increased.

But diagnosing more cases of cancer may not necessarily be a good thing. I know that sounds like a ludicrous proposition - surely if you have a breast cancer lurking you should know about it? Again, it is crucial to distinguish between breast screening - looking for breast cancer where there are no symptoms of having it - and ’symptomatic’ tests, done to investigate a complaint such as a lump or swelling. If there is a breast symptom like a lump, there is no debate: swift investigations are required. However, the rationale for investigating and treating some breast cancers picked up at breast screening where no symptoms were apparent can often be less clear.

As Wilson and Jungner wrote, it is no good finding more cancers if we cannot then improve the prognosis for the person. Finding cancer early by screening may contribute to a phenomena known as the ‘increase in lead time’ a well-known distorter of cancer survival statistics.

In other words, if more cancers are diagnosed earlier, it may look as though the survival is longer, but in fact, all you may be doing is increasing the length of time - the lead time - when you know about the cancer, and not actually treating the cancer any better. In other words, screening may put you on the train earlier, but the time of arrival at your destination may remain the same.

And then there is the risk associated with the screening process itself. In 2003, in a Review of Radiation Risks in Breast Screening (published on www.cancerscreening.nhs.uk) they concluded that the breast screening programme was justified in terms of radiation protection. They estimated that the risk of radiation causing cancer in a woman attending for mammographic screening was 1 in 20,000 per visit: and for every 170 cancers detected by the breast screening programme, one cancer is caused by it.

So exactly how many lives are saved by breast screening? It may be fewer than you think. In a Summary for the Evidence for the U.S. Preventive Services Task Force, published in the journal Annals of Internal Medicine in 2002, researchers analysed the latest data on breast screening. They found that they needed to screen over 1,200 women for over 10 years in order to save one life from breast cancer.

And in the process of screening all those women, many cancers will be picked up that are not, the traditional sense ‘cancer’. All cancers are not the same. Some are aggressive, some are not. Not all breast cancer picked up at screening will interfere with that woman’s lifespan. But, having found the cancer via screening, a woman will be offered a course of what may be painful and unpleasant treatment for a ‘cancer’ which may never have spread or have killed her.

Dr H Gilbert Welch, in his book, Should I Be Tested for Cancer? (University of California Press, 2004) calls this type of cancer ‘pseudodisease’. As Welch writes: “Because doctors can never be sure who has pseudodisease and who has a true disease, we tend to treat everybody. Patients with pseudodisease cannot benefit from treatment, however, because they were never going to develop symptoms of cancer in the first place. Instead they experience the emotional burden of a cancer diagnosis and face the side effects, complications, and risk of death associated with cancer therapies. And the way that patients get diagnosed with pseudodisease is by having a test looking for early cancer.”

Such early cancers - for example, ductal carcinoma in situ (DCIS) - make up around 20 PER CENT of breast cancers found at screening. The natural history of DCIS is uncertain. Some - the minority- will go on to become recognisably invasive cancers over the following decades. But most will not. A woman told she has DCIS is therefore placed in a difficult position does she have cancer or not? The distinction - even under a microscope - between ’cancer’ and ‘not-cancer’ is not always clear cut - especially in the realm of the borderline type of results thrown up by screening.

So should you sign up for every screening test going? There are no easy answers. There are actually only few diseases which are silent, yet treatable before they cause problems. For example, the ‘heelprick’ test, where newborn babies are screened for certain rare but potentially damaging diseases – e.g. congenital hypothyroidism and certain metabolic disease - is seen as being a successful screening test. These diseases in newborns can cause severe brain damage, but prompt treatment – diet change or medication – is a highly successful preventative intervention.

Nevertheless, excellent research continues to be done into areas where screening is very likely to have benefit, for example, detection of bowel cancer, or ultrasound screening of the aorta to look for potentially dangerous widening (aneurysm). The ideal is to prevent disease, but if this isn’t possible, then screening and treatment before disease causes harm is the next best thing. In the next few years, we may see targeted screening tests made available to those with higher genetic risk of particular disease, and the hope is that this would yield less false positive tests and more true positives – which, in turn, will bring different dilemmas to individuals of above-average genetic risk. In the meantime, screening may be helpful, but it is not a benign act.

Before operations, doctors ask patients to sign consent forms, and are obliged to outline the potential for risk as well as benefit. Surely we should hold the same standard for screening tests. Far from assuming that screening is universally beneficial, and universally desired, people should know the possible benefits, problems and uncertainties of screening before seeking request, not mere consent, for the test. And that process is not risk free.

Link to this article [subscription needed]