Specificity

Specificity measures how well a test recognizes people who are free of a condition. It is the number of true negatives (unaffected people who test negative) divided by everyone without the condition, which is the true negatives plus the false positives. A test with 95 percent specificity correctly clears 95 of every 100 unaffected people and wrongly flags the other 5, who receive a false-positive result. No real test is perfectly specific.

Specificity and sensitivity usually pull in opposite directions. Raising the threshold for calling a result positive cuts false alarms and raises specificity, but more true cases slip through. A highly specific test is valuable for confirming a diagnosis, summed up in the mnemonic SpPin: when specificity is high, a positive result helps rule the condition in. This matters because false positives can bring anxiety, further testing and unnecessary treatment.

Specificity should not be confused with negative predictive value, which tells a person how likely a negative result is to be correct and changes with how common the condition is. In biochemistry and immunology the same word means something different: how selectively an enzyme or antibody acts on its particular target.