Lead-time bias, as described in the passage, inflates measured survival because

screened patients receive more aggressive treatment than unscreened patients.

patients detected by screening tend to be younger than those diagnosed after symptoms.

survival is counted from diagnosis, so an earlier diagnosis lengthens it even if death comes at the same moment.

some screened patients are cured of cancers that would otherwise have killed them.

According to the passage, screening preferentially detects slow-growing tumours because such tumours

are more common than fast-growing ones in the populations invited to screen.

produce clearer images on the tests currently in use.

respond better to treatment once they have been identified.

spend a long period present in the body without producing symptoms.

It can be inferred that overdiagnosis makes a screening programme look effective because

patients with overdiagnosed tumours are excluded from the survival figures.

tumours that would never have caused illness are treated and then counted among the programme's successes.

overdiagnosed patients are more likely to attend subsequent screening rounds.

the treatment given for such tumours is unusually effective.

The passage indicates that population mortality is a better measure than survival chiefly because it

is unaffected by when a diagnosis is made or by how many harmless tumours are found.

can be established more quickly and at lower cost.

counts deaths from all causes rather than from the targeted cancer.

applies only to those who accept the invitation to be screened.

The passage suggests that the decision to narrow or withdraw a screening programme now rests on

the five-year survival rate among those the programme detects.

the proportion of the invited population that attends.

whether trials show a reduction in deaths from the targeted cancer against the treatment burden generated.

the availability of a more sensitive test for the same cancer.

←

Overdiagnosis and the Limits of Screening

PRACTICE

100:00+4 −1MCQ Single Answer

Reading Passage

The argument for cancer screening is easy to state and hard to resist. A tumour found early is smaller, more localized, and more likely to be curable than one found after it has produced symptoms. Programmes built on this reasoning have been adopted across the wealthy world, and the statistic most often cited in their support is striking: among people whose cancers are detected by screening, five-year survival is dramatically higher than among those diagnosed after symptoms appear. That statistic, however, cannot bear the weight placed on it, for two reasons. The first is that survival is measured from the moment of diagnosis. Finding a cancer three years earlier adds three years to measured survival even if the patient dies on precisely the day they would have died anyway — an artefact known as lead-time bias, which inflates the apparent benefit of any test that moves diagnosis earlier. The second reason is subtler. Screening does not sample tumours at random; it preferentially detects those growing slowly enough to be present without symptoms for a long time. Some of these would never have caused illness at all during the patient's life. Every such case is treated, and every one is then counted among the programme's successes, so a test that finds harmless tumours appears to save lives in exact proportion to how many harmless tumours it finds. The measure that resists both problems is mortality in the whole population invited to screen, compared against a population that was not. It is a far more demanding thing to establish, requiring large trials and long follow-up, and it has produced an uncomfortable picture: some programmes reduce deaths from the cancer they target by a clear margin, while others show little or no reduction while generating substantial treatment of disease that would never have declared itself. That distinction, rather than the survival figures, is what now determines which screening programmes are recommended and which have been narrowed or withdrawn.

The passage is primarily concerned with