What a screening test can and cannot tell a person

Two ways a test can be wrong

Every test can produce a positive result in someone without the condition and a negative result in someone who has it, and both errors occur in practice. The first kind leads to further investigation, anxiety and sometimes treatment for something that was not there, which is a real cost rather than an inconvenience. The second provides false reassurance, which can delay attention to symptoms that would otherwise have prompted a visit to a doctor.

Test design involves trading one type of error against the other, since adjusting a threshold to reduce one generally increases the other. This is why no test is simply accurate, and why the question is always accurate in which direction and at what cost.

Why rarity changes everything

When a condition is uncommon in the population being tested, most positive results come from the much larger group of people who do not have it. This holds even for tests that perform well, and it is one of the most counterintuitive facts about testing for people encountering it for the first time. It explains why screening a general population differs so much from testing someone who already has symptoms suggesting a particular condition.

The prior probability, meaning how likely the condition was before the test, does most of the work in interpreting what the result means. A clinician bringing that context to a result is doing something an isolated number on a report cannot do by itself.

Overdiagnosis as a distinct problem

Some conditions detected by screening would never have caused symptoms or shortened life, and finding them converts a healthy person into a patient. This is not a false positive, since the condition genuinely exists, and that is precisely what makes it difficult to explain and difficult to avoid. The harm comes from the investigation and treatment that follows, both of which carry their own risks and their own consequences.

How much overdiagnosis occurs in any given screening programme is contested and varies substantially between conditions and between approaches. It is one of the main reasons expert bodies sometimes recommend against screening that intuitively seems obviously worthwhile.

Why programmes have eligibility criteria

Organised screening programmes specify ages and risk groups because the balance of benefit and harm shifts with how common the condition is in that group. Testing outside those criteria is not automatically better, and it can tip the balance towards the harms without adding much detection. This is genuinely counterintuitive, since more information feels like it should always be preferable to less.

The criteria are set by reviewing accumulated evidence and are revised as that evidence changes, which is why they differ between countries. Anyone uncertain whether a particular test is appropriate for them should raise it with a doctor rather than deciding from general reading.

What to do with an actual result

A result outside a reference range is a prompt for further conversation rather than a diagnosis, and many such results have ordinary explanations. Reference ranges themselves are constructed from measurements in a reference population, which means a proportion of healthy people fall outside them by definition. Direct-to-consumer testing produces results without the clinical context that makes them interpretable, which is a recognised source of unnecessary alarm.

Nothing in this description helps interpret any specific result, because interpretation depends on the individual and their history. The appropriate response to an unexpected result is a conversation with a clinician rather than a search for reassurance elsewhere.