Sensitivity and specificity
Sensitivity is the proportion of people who have a condition whom the test correctly identifies as positive. Specificity is the proportion of people without the condition whom the test correctly identifies as negative.
- Sensitivity = true positives ÷ (true positives + false negatives)
- Specificity = true negatives ÷ (true negatives + false positives)
A test with high sensitivity misses few cases, so it gives few false negatives. A test with high specificity gives few false positives. The two values can trade against each other: raising or lowering the cut-off point of a test changes both.
Predictive values and prevalence
Patients usually ask a different question: if my result is positive, how likely is it that I have the condition? This is the positive predictive value (PPV). The negative predictive value (NPV) is the probability that a person with a negative result truly does not have the condition. Predictive values depend on how common the condition is among the people tested.
As a worked example, suppose a test with 99% sensitivity and 99% specificity is used in 10 000 people, of whom 1% (100 people) have the condition. The test detects 99 of the 100 true cases, but it also gives a positive result in 1% of the 9 900 unaffected people, which is 99 false positives. Of the 198 positive results, only half are true positives. Used in a group where the condition is more common, the same test would have a higher PPV.
Why false results happen
Timing is a common cause, because early in an infection there may not yet be enough of the target for the test to detect. For HIV, the CDC describes a window period after exposure during which a test may not yet detect infection: about 10 to 33 days for nucleic acid tests, 18 to 45 days for laboratory antigen and antibody tests, and 23 to 90 days for most antibody tests. A test taken too early can therefore be falsely negative.
Biology can also produce misleading results: in prenatal cell-free DNA screening, the DNA tested comes mainly from the placenta, and a chromosome change confined to the placenta can produce a false positive result for the fetus. Problems at any step from sample collection to reporting can also affect a result, which is why laboratories use validated methods, controls and quality assurance.
Retesting and confirmatory testing
A result should be interpreted together with the person's symptoms, history and risk. When a result does not fit the clinical picture, the clinician may discuss it with the laboratory or repeat the test. After a possible HIV exposure, a negative result should be repeated once the window period for that test type has passed, and a positive HIV antibody result needs a follow-up test to confirm it. In prenatal screening, the Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZCOG) recommends that anyone with a high chance result on a screening test is offered diagnostic confirmation before further management decisions are made.
Screening tests and diagnostic tests
A screening test sorts people into higher and lower chance groups. A diagnostic test aims to establish whether a condition is present. A positive screening result is a reason for further assessment, not a diagnosis, and a negative screening result lowers the chance of a condition without excluding it entirely.