Open Medicine EU

More experts are questioning the practice of screening in particular cases – the testing of apparently healthy populations for underlying risk factors or undiagnosed conditions, such as some cancers, with a view to prevention or early treatment. Supporters of screening sometimes respond to their critics with more heat than light, but I don’t know who is right.

At first sight, screening may seem to be a sensible precaution, so why question it at all? Well, a lot depends on the prevalence of the risk factor or condition, the reliability of screening tests and the effects of early treatment. These are factors that need good and robust research.

An epidemiologist friend once explained to me some of the maths of screening, and the results were surprising. To take an (unoriginal) example, suppose one in fifty people (2%) are likely to have a particular risk factor or condition. No tests are perfect and suppose the screening test will miss an underlying condition in 10% of cases (false negative) and give a false positive in another 10% of cases. Finally, and keeping the numbers simple, suppose that we screen 100,000 people.

In this example, 2000 people will have the risk factor or condition in question. Since the test is 90% effective the screening will miss 200 of these. On the other hand, 98 000 people will not have the risk factor or condition. However, the screening test will throw up a false positive in 10% of these cases or 9 800 people.

So we have:
1800 people correctly detected,
200 wrongly told their test was negative, and
9800 wrongly told their test was positive.

For the 9800 people with false positives the next stage may be relatively easy or highly intrusive, depending on circumstances. It will often carry some risk. A 2% risk in the treatment would translate into adverse effects for nearly 2000 people who had no problem before entering screening.

What of the 1800 people correctly detected? How effective will the prescribed treatment be and how will it affect wellbeing? There is much discussion about the efficacy of particular treatments, following early detection. Some people with risk factors never develop the disease or condition. Few if any treatments are 100% effective. Some treatments can be effective against a particular risk but may be miserable for the patient. Some treatments may ward off the threatened condition but provoke other problems. (A treatment that is good at warding off a cancer may increase the rate of heart attacks, for example.)

A recent article in the Economist cites research suggesting that for every life saved by a type of testing for prostate cancer, “the treatment inspired by those tests would cause one man to develop a serious blood clot, two to have heart attacks and at least 30 to become impotent or incontinent”.

I am not arguing against screening but for better information and understanding in each case, for patients and policy makers – about the incidence of the problem, the reliability of the tests, the outcomes for false positives and negatives, plus the true efficacy and survival rates of early treatment. These are the kind of factors that must be considered when assessing if screening is worthwhile.

Screening may seem to be effective only for some groups – the over-fifties, for example. In this case younger people may be mistaken in arguing that they too should have a “right” to screening – in their case screening may do more harm than good. Everything depends on the circumstances of each case.

Pressure for screening can build up among advocacy groups, the media, professionals, makers of screening tests and for-profit providers with their PR and advertising allies. Advertising for some screening services can be downright misleading, exploiting anxiety and falsely promising “peace of mind”. The demand for screening, and more screening, may seem reasonable, but the truth can be counter-intuitive – hence the need for sound evidence and good science.

(Finally, a caveat: screening is the testing of people with no apparent symptoms; people with symptoms should seek appropriate advice.)

Author :