Contents
Academy · Research Data, Analytics & AI · Sensitivity and specificity
Sensitivity and Specificity: Calculation, Trade-Offs and Distinction from Predictive Values
A test calibrated to detect every true case will also yield more false positives, and a test calibrated to avoid false positives will miss more true cases; sensitivity and specificity quantify each of these errors.
Sensitivity is the proportion of people who truly have a condition who test positive for it. Specificity is the proportion of people who truly do not have the condition who test negative for it. Both are properties of the test itself, calculated against a reference standard, and neither alone describes how useful a positive or negative result is for an individual patient.
Learning objectives
- How sensitivity and specificity are calculated from a 2x2 table against a reference standard.
- Why the two typically trade off against each other as a test's positivity threshold shifts.
- Why sensitivity and specificity are not the same as predictive values, and when each matters more.
Common misconceptions
- “A test with 95% sensitivity means 95% of positive results are correct.” Sensitivity is calculated among people who truly have the condition (true positives divided by true positives plus false negatives); it provides no direct information about the proportion of positive results that are correct, which is instead the positive predictive value, a related but distinct quantity that also depends on the prevalence of the condition in the population tested.
- “A more sensitive test is simply a better test.” Trevethan notes that sensitivity and specificity typically trade off against each other as a test’s decision threshold is adjusted: a test made more sensitive (to detect more true cases) usually becomes less specific (producing more false positives), so which test is “better” depends on which error is more costly in the specific clinical context.
Calculation of sensitivity and specificity
Trevethan sets out the standard 2x2 framework against a reference standard: sensitivity is true positives divided by (true positives plus false negatives), the proportion of truly affected individuals correctly identified; specificity is true negatives divided by (true negatives plus false positives), the proportion of truly unaffected individuals correctly classified as negative1. Both figures depend entirely on comparison against an accepted reference standard that establishes true disease status independently of the test being evaluated.
Trade-off between sensitivity and specificity across thresholds
For tests with a continuous or graded result (such as a biomarker level), moving the positivity threshold to detect more true cases (raising sensitivity) almost always also classifies more people without the condition as positive (lowering specificity), and vice versa1. This trade-off is the basis of the ROC curve, which plots sensitivity against 1 minus specificity across every possible threshold, allowing a threshold to be chosen deliberately according to which error (missed cases or false positives) is more consequential in the intended setting.
Distinction from predictive values
Sensitivity and specificity are properties of the test, independent of the prevalence of the condition in the population being tested. Positive and negative predictive values (the proportion of positive results that are true positives, and the proportion of negative results that are true negatives) depend additionally on the condition’s prevalence in that population1. The same test, with identical sensitivity and specificity, will have a markedly different positive predictive value in a high-prevalence population (such as a specialist referral clinic) than in a low-prevalence population (such as general community screening), which is why sensitivity and specificity alone cannot answer the patient’s question of how likely it is that a positive result indicates the condition.
Key points
Key points
Sensitivity and specificity are properties of a test calculated against a reference standard; they typically trade off against each other as the positivity threshold shifts, and neither is equivalent to a predictive value, which additionally depends on the prevalence of the condition in the population being tested.
Practical application
Practical application
- Sensitivity and specificity should be reported against a clearly stated reference standard. The reference standard should be named explicitly and the full 2x2 cross-tabulation reported, in accordance with STARD, rather than summary statistics alone.
- The threshold used should be stated if the test produces a graded or continuous result. The specific cut-off applied should be reported and, where relevant, the full receiver operating characteristic (ROC) curve should be presented or referenced, rather than a single threshold in isolation.
- Predictive values should be reported separately, with the population prevalence stated. Sensitivity and specificity should not be presented as though they answer the predictive-value question; positive and negative predictive values (PPV and NPV) should be calculated and reported separately, with the assumed or observed prevalence stated explicitly.
Frequently asked questions
The choice depends on the clinical consequence of each type of error: a screening test for a serious, treatable disease often prioritises sensitivity to avoid missed cases, whereas a confirmatory test used before a risky treatment often prioritises specificity to avoid unnecessary treatment.
Yes. A genuinely strong test can have high values of both, but this reflects the test’s underlying discriminative ability and cannot be achieved simply by adjusting the threshold; for a fixed test, threshold adjustment alone trades one against the other.
Sensitivity and specificity themselves are relatively stable properties of the test, but the practical usefulness of a positive or negative result (its predictive value) shifts with the prevalence of the condition in that specific setting.
References
- Trevethan R. Sensitivity, Specificity, and Predictive Values: Foundations, Pliabilities, and Pitfalls in Research and Practice. Frontiers in Public Health. 2017;5:307. read 10 October 2026. View on PubMed Central
Every source above is a peer-reviewed article, a publisher’s record of one, or a standard issued by the body that owns it. Each carries the date it was last read. Where a standard moves, this article is revised rather than left up with a note.
How to cite this article
Pick your referencing style and copy it. Your university or target journal may vary the punctuation — check its guide if it has one.
Pubrica Academy. (2026, October 10). Sensitivity and Specificity: Calculation, Trade-Offs and Distinction from Predictive Values. Pubrica. https://pubrica.com/academy/concepts-definitions/sensitivity-and-specificity/
Pubrica Academy. “Sensitivity and Specificity: Calculation, Trade-Offs and Distinction from Predictive Values.” Pubrica, 10 Oct. 2026, pubrica.com/academy/concepts-definitions/sensitivity-and-specificity/.
Pubrica Academy (2026) Sensitivity and Specificity: Calculation, Trade-Offs and Distinction from Predictive Values. Pubrica. Available at: https://pubrica.com/academy/concepts-definitions/sensitivity-and-specificity/ (Accessed: 15 October 2026).
Pubrica Academy. “Sensitivity and Specificity: Calculation, Trade-Offs and Distinction from Predictive Values.” Pubrica. October 10, 2026. https://pubrica.com/academy/concepts-definitions/sensitivity-and-specificity/
Pubrica Academy, ‘Sensitivity and Specificity: Calculation, Trade-Offs and Distinction from Predictive Values’ (Pubrica, 10 October 2026) <https://pubrica.com/academy/concepts-definitions/sensitivity-and-specificity/> accessed 15 October 2026.
Pubrica Academy. Sensitivity and Specificity: Calculation, Trade-Offs and Distinction from Predictive Values. [Internet]. Pubrica; 2026 Oct 10 [cited 2026 Oct 15]. Available from: https://pubrica.com/academy/concepts-definitions/sensitivity-and-specificity/
Pubrica Academy, “Sensitivity and Specificity: Calculation, Trade-Offs and Distinction from Predictive Values,” Pubrica, Oct. 10, 2026. [Online]. Available: https://pubrica.com/academy/concepts-definitions/sensitivity-and-specificity/
@misc{sensitivity_and_specificity_2026,
author = {{Pubrica Academy}},
title = {Sensitivity and Specificity: Calculation, Trade-Offs and Distinction from Predictive Values},
howpublished = {Pubrica},
year = {2026},
month = {oct},
url = {https://pubrica.com/academy/concepts-definitions/sensitivity-and-specificity/},
note = {Accessed: <span data-cite-acc="iso">2026-10-15</span>}
}This article is written to be learned from. Cite it if you are referring to what it says; if you are using the evidence behind it, cite the sources listed above instead.
Diagnostic-accuracy reporting check
This checklist applies when reporting a diagnostic test’s performance, alongside STARD.
[ ] Reference standard named explicitly [ ] Full 2x2 cross-tabulation reported [ ] Sensitivity and specificity reported with confidence intervals [ ] Positivity threshold stated, if the test is graded/continuous [ ] Predictive values reported separately, with prevalence stated [ ] ROC curve presented or referenced where a threshold was chosen from a range
Pubrica AcademyWritten and maintained by Pubrica’s subject matter experts. Articles are revised when the underlying standards change, and each article states the date on which it was last reviewed.
How Pubrica can help
Pubrica assists in structuring diagnostic-accuracy results in accordance with STARD, including the full 2x2 cross-tabulation and a clear distinction between sensitivity and specificity on the one hand and predictive values on the other.