Specificity

Specificity is the proportion of people who truly do not have a condition that a test correctly identifies as negative. It answers the question "among those without the disease, how many does the test clear?" and is the counterpart to sensitivity among the two intrinsic accuracy measures used to evaluate diagnostic and screening tests against a reference standard.

Definition

Specificity is the conditional probability that a test result is negative given that the disease is truly absent, calculated as the number of true negatives divided by the total number of people without the disease (true negatives plus false positives).

Scope

This entry defines specificity as the true-negative rate, places it within the 2x2 table of test result against true disease status, explains its complement (the false-positive rate), and notes how it is estimated, how it can vary with the patient spectrum, and how it contributes to likelihood ratios. It is a methodological topic and does not advise on the use of any particular test.

Key concepts

• True-negative rate
• False-positive rate (1 - specificity)
• Reference (gold) standard
• Conditional probability given no disease
• Spectrum bias
• Negative likelihood ratio

Mechanisms

Specificity is computed down the non-diseased column of the 2x2 table: of all subjects whose true status is disease-free, it is the fraction the test correctly calls negative. Because it conditions on true non-disease status, specificity is in principle independent of disease frequency and so characterises the test rather than the population. Its complement, one minus specificity, is the false-positive rate — the proportion of healthy people wrongly flagged. A highly specific test, when positive, helps rule a condition in, because few disease-free people produce positive results. Specificity combines with sensitivity to form the negative likelihood ratio and, through the false-positive rate, defines the horizontal axis of the ROC curve. Like sensitivity, measured specificity can depend on the spectrum of non-diseased subjects studied, since co-existing conditions can raise false positives.

Clinical relevance

Specificity is a standard yardstick for how well a test avoids labelling healthy people as diseased and is emphasised where false positives carry significant cost, such as unnecessary follow-up after screening. The concept supports critical appraisal of diagnostic evidence; it describes a property of a test and is not a basis for individual diagnostic or treatment decisions.

Epidemiology

In population screening, even modest shortfalls in specificity matter because most people tested are disease-free, so a small false-positive rate applied to a large healthy majority can generate many false alarms. This interaction of specificity with prevalence is a central theme in appraising whether a screening programme produces acceptable harms relative to benefits.

History

Specificity entered medical statistics paired with sensitivity from the theory of classification and signal detection and was popularised for clinical readers through expository statistical writing in the 1990s. Methodological work in the 1970s highlighted how the spectrum of non-diseased subjects could bias the measure.

Debates

Why does specificity dominate the false-positive problem in screening?

Because the non-diseased usually vastly outnumber the diseased in screening, even a high specificity can leave an absolute number of false positives that strains follow-up resources and exposes healthy people to harm.

Key figures

• Douglas Altman
• Martin Bland
• Jonathan Deeks
• David Grimes
• Kenneth Schulz

Seminal works

• altman-bland-1994a
• ransohoff-feinstein-1978
• deeks-altman-2004

Frequently asked questions

Does a highly specific test rule disease in when positive?

A high specificity means few healthy people test positive, so a positive result is more convincing for disease; the full interpretation still depends on sensitivity and the underlying prevalence.

What is the false-positive rate?

It is one minus specificity: the proportion of people without the condition who are wrongly classified as positive by the test.