three adults peering down microscopes

Diagnosis of whooping cough

Tests used in the diagnosis of whooping cough (pertussis).

Sometimes it is acceptable to diagnose whooping cough using the WHO clinical definition which is three weeks of paroxysmal coughing. This a very poor way of diagnosing whoopingcough-pertussis because other infections can cause a paroxysmal cough, and pertussis does not always cause these precise symptoms but can cause just an ordinary cough or be asymptomatic.

There are 3 different tests. Culture, antibody detection, and PCR are all used in the diagnosis of whooping cough.

PCR is good in the first 3 weeks. Antibody tests are good after 2 weeks. Culture is good in the first 3 weeks but only with meticulous technique.

Which test is done may depend where you live.

In many developed countries a PCR test on a throat or nasal swab is now standard (in Australia and the USA for example, and now available in UK primary care). In many other countries antibody tests on a blood sample is normal in adults and oral fluid antibody tests may be done in children. In many countries the test that is done will depend on the laboratory that is used. 

More detail

Antibody testing in the diagnosis of whooping cough 

It is common but being replaced by PCR.

A blood sample taken after a minimum of two weeks of illness is used. By measuring IgG antibodies to pertussis toxin it is possible to say whether it is likely the patient has had pertussis infection with 90% accuracy, provided there has been no pertussis immunization in the previous 12 months.

This antibody is usually measured as International Units (IU), and a level over 70 IU can be taken as very strong evidence of recent infection. Different countries may use different thresholds from 70 IU. IgA is sometimes measured instead, or sometimes both. IgA only rises after natural infection. IgG rises after either natural infection or immunization.

The test will be falsely negative in 10% of pertussis infections. It will also be negative in Bordetella parapertussis and Bordetella holmesii infections, (which cause similar symptoms). That is because they do not produce pertussis toxin, so test negative.

Oral fluid obtained by using a special sponge kit can be tested for pertussis toxin antibodies in the same way. It it not quite as accurate as blood testing. There are more false negatives. Oral fluid testing is usually reserved for children because of the difficulty of getting blood from them.

Here is a reference to a relevant European document on single sample serological diagnosis It opens in a new Tab

Antibody tests can be done late in the illness and still show positive which is a big advantage. 

In the United Kingdom a blood specimen from suspected cases should be sent to the local NHS laboratory requesting ‘pertussis antibodies’. Results are obtained in 1-2 weeks. It can be difficult to persuade doctors to do the test. In the UK there are clear guidelines that include testing any patient with a paroxysmal cough of more than 2 weeks duration. There are other circumstances described and the actions to be taken. 

UK guidelines for doctors here

Drawing these guidelines to the attention of your doctor may sometimes be necessary as very few will be familiar them (nobody can possible remember them all!). 

In the USA there is less likelihood that a doctor will refer to CDC guidelines as state health practices may predominate, and they are sometimes a bit out of date. There is a CDC website page you might find useful.

PCR (polymerase chain reaction)

This is  a more successful way of detecting the organism. It is best done in the first three weeks of symptoms. Generally the earlier the better. It detects its unique DNA pattern. This involves getting secretions from the back of the nose or throat by swab or aspiration and testing in a specialist laboratory.  A result can be obtained in 24 to 48 hours.

A negative PCR does not rule out pertussis especially if taken in the later stages. It should be positive right from the first day of the illness and is reliable for 3 weeks and may remain positive for 4 weeks or more.

The PCR test depends on traces of the organism being present, alive or dead. Since it detects minute quantities of genetic material it is more likely to be positive than culture, and for a longer period of time.

PCR has the advantage that it can be successful on a throat swab, unlike culture that has to be taken from an area of ciliated epithelium where the bacteria are living which is at the back of the nose. A throat swab for PCR should be sent to the lab dry, not in transport medium, although that does not usually stop it being tested.

One thing that can happen with PCR testing that can be confusing is that it detects infections that may not be associated with illness from whooping cough. Some people get the infection and get no significant symptoms, or mild symptoms but they will be PCR positive.

PCR can be too sensitive

This can be a problem for statistics. For example, if a parent takes a child with whooping cough to a doctor and a sample is taken for PCR, the parent and doctor may arrange for other children in contact to be tested too, even if they have no symptoms. Some may show PCR positive but not go on to develop whooping cough.

A positive PCR from such cases will show up in the pertussis statistics and make the incidence look greater. Previous to PCR availability, only clinical whooping cough, blood testing and culture was counted for statistical purposes. These three are a good measure of clinical whooping cough. PCR, contrastingly, measures pertussis infection, which can be quite different, as many infections do not turn into whooping cough. 

If comparisons are to have any validity, clinical whooping cough needs to be recorded and notified separately for PCR positives.

This may explain some of the resurgence described in Australia. That country is heavily reliant on PCR.

A per-nasal swab for Bordetella pertussis detection
A per-nasal swab for bacterial culture of B. pertussis


The oldest and most difficult way is to try to culture the causative organism (Bordetella pertussis) from the back of the nose. This involves passing a swab on a wire through a nostril to the back of the throat and sending it to a medical lab. This may take 5 to 7 days.  If Bordetella pertussis or parapertussis grows, this is proof that it is whooping cough. Parapertussis also causes whooping cough. It is much less common, possibly 1 in 100 cases. It may be less severe because it does not produce pertussis toxin. Culture by per-nasal swab only identifies about a third of cases, even in the best hands.

Unfortunately the organisms is delicate, killed easily by many antibiotics and has often been eliminated from the body by natural defences by the time the diagnosis is suspected. It is easiest to find in the first 2 weeks, but very unlikely after 3 weeks. The patient has often had it for 3 weeks before whooping cough is suspected, so it is unusual to get a positive culture in whooping cough. In other words, if a swab is negative, you can still have whooping cough.

In practice the diagnosis has often to be made on symptoms and course of the illness alone, unless blood or oral fluid antibody tests or PCR can be done.  

Todar’s online bacteriology chapter on pertussis


This page has been reviewed and updated on 20 March 2023