FolderUnderstanding and preventing

Diagnostic tests: understand PCR, antigen and serology

A test answers a specific question about an agent, a sample and a moment. Its value depends on the method, the quality of the sample and the probability of infection before the test.

Topic updated: 27/10/2026

AI-generated illustration, not evidence of a health event. Image generated by artificial intelligence. IA-generated fictional scene; it does not represent an identified patient or clinical evidence.

EssentialShort answer

To be retained

The starting points

PCR detects the pathogen’s genetic material, an antigen test detects a protein, and serology looks for an antibody response. A negative result cannot always rule out infection, particularly with an early sample or a less sensitive test. Selection and interpretation must follow guidance specific to the pathogen being tested for.

Deepen ▼

01Reference

Three methods, three questions

The following principles are illustrated by respiratory viral infections; indications change according to the disease.

Molecular or PCR test

Look for a genetic sequence of the agent in a suitable sample. Detection may persist after the most contagious phase for some viruses.

Antigenic test

Look for a protein agent. It can give a quick result but more easily lacks an infection when there is little detectable material.

Serology

Search for antibodies produced by the person. It can provide information on past exposure, but is not generally suitable for early diagnosis.

Combined test

Some devices look for several agents in the same sample; the exact target list and performance are listed in the package leaflet.

02Reference

Why a result can be wrong

A negative or positive value does not have the same meaning in all contexts.

Time of sampling

A test carried out too early may precede the appearance of a detectable amount of the agent or antibody response.

Quality of sampling

The site, technique, transport and conservation influence what the laboratory or device can detect.

Test performance

Sensitivity describes the ability to detect infections; specificity, the ability to limit positive results wrongly.

Probability before testing

Symptoms, contacts and circulation of the agent change the probability that a result will be the actual situation.

03Reference

Read the report usefully

Before drawing a conclusion, identify the question that the test was to answer.

Agent and sample

Check which agent was wanted and on what type of sampling; a result does not exclude untested agents.

Date and symptoms

Note the date of possible exposure, onset of symptoms, and sampling.

Negative result

If suspicion remains strong, ask if another sample or repeat is indicated for this specific test.

Positive result

Follow the applicable instructions and discuss care with a professional, especially in case of a risk factor.

In practiceObserve and understand

Why ‘99% reliable’ is not enough

Sensitivity describes detection among people with the condition; specificity describes negative results among those without it. Positive predictive value asks a different question: among positive results, how many correspond to the condition?

Why ‘99% reliable’ is not enough — explanations and limits
Concept	Denominator	Do not confuse with
Sensitivity	People with the condition being tested for.	The proportion of true positives among all positive tests.
Specificity	People without this condition.	The absence of false negatives.
Positive predictive value	All positive results.	A fixed test property independent of the population.

Explore results in a hypothetical population

Example with 10,000 people: 1% have the condition, with 90% sensitivity and 99% specificity. Change one parameter to explore the denominator’s role.

Expected model counts per 10,000 people
Test result	Condition present	Condition absent
Positive	90 true positives	99 false positives
Negative	10 false negatives	9 801 true negatives

47.6 %Positive predictive value

99.9 %Negative predictive value

Of 189 expected positive results, 90 correspond to the condition in this hypothetical example.

Understand the calculations

True positives = population × prevalence × sensitivity. False positives = population × (1 − prevalence) × (1 − specificity). Positive predictive value = true positives / all positives. Negative predictive value = true negatives / all negatives. Convert percentages to proportions for calculation. Displayed counts are rounded; proportions are calculated before rounding.

Keep in mind

The selected values are hypothetical. The model assumes constant performance and a condition known through a perfect reference. It does not include sample quality, timing or uncertainty and cannot interpret your personal result.

References for these explanations

CDC — Evaluating surveillance systems: sensitivity and predictive value ↗

Sources consulted on 14/10/2026. Examples and diagrams are educational; the original guidance defines its scope.

QuestionsClarifications

Frequently Asked Questions

Does a positive PCR always mean that the person is contagious?

No. For some agents, genetic material remains detectable after the period of contaginess. The method and the clinical context matter.

Does a negative self-test ignore symptoms?

No. A negative antigenic test may miss an infection and only look for the agents indicated on its package leaflet. See the test-specific repeat recommendations.

Does a positive serology prove a current infection?

Usually not. Antibodies may reflect a previous infection or, depending on the test, vaccination. Their interpretation is specific to the disease.

Should all infections be tested the same way?

No. The choice of sampling and method depends on the agent, the time frame, the clinical situation and the current recommendations.

Moving from reading to exchanging

Keep the results complete with their dates and prepare questions on their interpretation. Organize information for a health professional →

Institutional sources

External sources can evolve. Check their date and instructions for your territory.

PratiqueCheck before deciding

Positive or negative: identify the question

Reading guidance and interpretation limits
Concept or situation	What it describes	Point to watch
Pathogen detection	A test detects a specific pathogen or one of its components.	Sample, testing time and method sensitivity.
Serology	Detects an immune response.	It does not always establish an active infection or lasting protection.
Screening	Looks for an infection in a defined context.	A result may require confirmation or a later follow-up test.