[Durgical](https://durgical.com/en/) › [DurgiPedia](https://durgical.com/en/durgipedia/) › [Epidemiology](https://durgical.com/en/epidemiology/) › [Diagnostic tests](https://durgical.com/en/epidemiology/areas/diagnostic/) › Reference standard (gold standard)

# Reference standard (gold standard)

- Concept

The best available way to know who has the disease; every other test's performance is measured against it.

Updated Nov 5, 2026

## Key points

- Every sensitivity and specificity is relative to the reference standard used.
- An imperfect reference distorts the performance of the test being evaluated.
- Incorporation bias: the test being evaluated is part of the reference definition, inflating its performance.
- Verification bias: the reference is applied only to people who test positive on the new test.
- The reference is read without knowing the result of the test being evaluated (blinding).

## What it is

The reference standard (gold standard) is the method accepted as the best available classification of diseased and healthy people: a biopsy, a culture, clinical follow-up or a combination of criteria. Every sensitivity and specificity is calculated against it, so its errors pass on to the test being evaluated. If the reference is imperfect, a new test can look worse (or better) than it really is. It must also be applied to everyone and independently of the new test's result.

## Worked example

- Data: 200 people, 100 truly diseased and 100 healthy. The reference detects 90 of the 100 diseased and gives no false positives.
- A new, perfect test is positive in all 100 diseased and negative in all 100 healthy.
- Against the reference: 90 diseased, all positive on the new test; apparent sensitivity = 90 / 90 = 100%.
- The 110 reference negatives include 10 undetected diseased people, who are positive on the new test.
- Apparent specificity = 100 / 110 ≈ 90.9%: those 10 correct results are counted as false positives.
- Reading: with an imperfect reference, a perfect test appears to have 9% false positives.

## Problems with the reference standard

| Problem | What happens | Usual effect |
| --- | --- | --- |
| Imperfect reference | The reference misclassifies some people | The new test's performance is distorted |
| Incorporation bias | The new test is part of the reference | Inflated sensitivity and specificity |
| Verification bias | Only some people get the reference, depending on the new test's result | Inflated sensitivity, reduced specificity |
| Unblinded reading | Whoever reads the reference knows the new test's result | Artificially high agreement |

## Markers

- gold standard
- reference

## In practice

- When no single reference exists, an expert panel, composite criteria or patient follow-up over time is used.
- A new test more sensitive than the reference can seem to have false positives that are really cases the reference misses.

## Common pitfalls

- Treating the reference as infallible.
- Including the evaluated test's result in the final diagnosis.
- Applying the reference only to positives.

## Memory aid

The ruler can be crooked too: the reference standard's errors are billed to the new test.

## Test yourself

Think of the answer, then tap the question to see it.

## What is incorporation bias?

Bias that arises when the evaluated test is part of the reference definition; its sensitivity and specificity come out inflated.

## A new test detects cases the reference misses. How do those cases appear in the study?

As false positives, which makes the new test's specificity look lower than it is.

## Why must the reference be applied to every participant?

Because applying it only to people positive on the new test causes verification bias.

[Add to my cards (3)](https://durgical.app/estudiar/?add=/en/epidemiology/reference-standard/)

## Connections

## Related 5

- [Verification (work-up) bias](https://durgical.com/en/epidemiology/verification-bias/)
- [Spectrum bias](https://durgical.com/en/epidemiology/spectrum-bias/)
- [Sensitivity](https://durgical.com/en/epidemiology/sensitivity/)
- [Specificity](https://durgical.com/en/epidemiology/specificity/)
- [Validity and reliability](https://durgical.com/en/epidemiology/validity-and-reliability/)

[← All of epidemiology](https://durgical.com/en/epidemiology/)

Educational content on epidemiological and statistical methods; the examples use made-up numbers to teach the calculation. Not a clinical or legal guide: current Colombian rules are in the official sources.

Sources: [Statistical Guidance on Reporting Results from Studies Evaluating Diagnostic Tests - Guidance for Industry and FDA Staff](https://www.fda.gov/regulatory-information/search-fda-guidance-documents/statistical-guidance-reporting-results-studies-evaluating-diagnostic-tests-guidance-industry-and-fda) (U.S. Food and Drug Administration, 2007); [Measures of Diagnostic Accuracy: Basic Definitions](https://europepmc.org/article/PMC/PMC4975285/) (EJIFCC, 2009).

## **Spotted a mistake?** We are always improving for the community. Tell us what is wrong and our team will review it. Report a mistake

Representative model, for educational purposes only.

---
Source: https://durgical.com/en/epidemiology/reference-standard/
License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Cite as: DurgiPedia (durgical.com), https://durgical.com/en/epidemiology/reference-standard/