Close help

How to use this Learning Station

On this Learning Station, you can read and test your knowledge. Tap on a book to open its chapter list. In each chapter, you can take a quiz to test your knowledge.

To take tests, you must register with your email address or cell number. It is free to register and to take tests.

For help email or call +27 76 657 0353.

Printed books

Learning is easiest with printed books. To order printed books, email or call +27 76 657 0353.

Visit for information.


Diagnosis of childhood tuberculosis

Take the chapter test before and after you read this chapter.

Open chapter test

Close test

First time? Register for free. Just enter your email or cell number and create a password.

Close test



When you have completed this unit you should be able to:

Confirming the clinical diagnosis of tuberculosis

3-1 How is the clinical diagnosis of tuberculosis confirmed?

A history of chronic cough and contact with an adult with tuberculosis must always suggest tuberculosis. However the suspected clinical diagnosis of tuberculosis is often difficult to prove, especially in children. Therefore special investigations are important to help confirm or reject the clinical diagnosis.

Special investigations are important to confirm the clinical diagnosis of tuberculosis.

3-2 Which special investigations are important in diagnosing tuberculosis?

A number of special investigations are useful in confirming a clinically suspected diagnosis of tuberculosis, including:

These special investigations must always be used together with a careful history and full physical examination. The tuberculin skin test and chest X-ray are particularly important.

The diagnosis of tuberculosis in children usually depends on a careful history, clinical examination, tuberculin skin test and chest X-ray.

Tuberculin skin tests

3-3 What is a tuberculin skin test?

This is a skin test done with tuberculin which contains protein from dead TB bacilli. Usually PPD (i.e. Purified Protein Derivative) in the form of tuberculin is used. It is safe as it does not contain live TB bacilli. The most accurate method of tuberculin skin testing is the Mantoux test, where a small amount of standardised PPD is injected into the skin (intradermally).

A tuberculin skin test determines whether the person has become infected with TB bacilli. Tuberculin skin tests usually become positive four to eight weeks after infection with TB bacilli.

Another less accurate method of skin testing is the Tine test.

The Mantoux skin test is the preferred method of tuberculin skin testing.

3-4 How is the Mantoux skin test done?

A 1 ml syringe and size 26 needle are used to inject 0.1 ml of tuberculin (standardised PPD) into the skin over the inner side of the left forearm. It is very important that the tuberculin is injected into the skin (intradermally) and not under the skin (subcutaneously). If the tuberculin is correctly injected into the skin, a raised, pale wheal of 5 mm is formed. This is slightly painful. If no wheal is raised, the tuberculin has been injected too deep in error. Incorrect injection under the skin may make the result difficult to interpret.

3-5 How should you read a Mantoux skin test?

The Mantoux skin test must be read by examining the site of the test two to three days (48 to 72 hours) after it is done. The widest transverse diameter (across the arm) of induration (raised, swollen, thickened area of skin) is measured. It is important that the induration and not the area of redness is measured. The diameter of the induration is best measured with a ruler. Never guess the size of a Mantoux skin test. The result should be reported in millimeters and not simply as positive or negative.

PPD stimulates a delayed reaction with sensitised lymphocytes being attracted to the site where they release inflammatory mediators which result in oedema (swelling) and erythema (redness).

3-6 How is the result of a Mantoux skin test interpreted?

The Mantoux skin test is regarded as positive:

  1. In immunosuppressed children (including HIV-infected children and children with severe acute malnutrition) if the transverse diameter of induration is greater than 5 mm.
  2. In all other children, whether or not they received BCG immunisation, if the transverse diameter of induration is 10 mm or more.

The greater the diameter the more significant is the result.

A Mantoux skin test of less than 5 mm in an immunosuppressed child or less than 10 mm in all other children is regarded as negative. However, both false-negative and false-positive Mantoux results may occur.

Both HIV infection and severe acute malnutrition reduce the response to PPD and may result in a false-negative Mantoux skin test.

3-7 What is the meaning of a positive skin test?

A positive skin test usually indicates that the child has been infected with TB bacilli. However it does not necessarily imply that the child has tuberculosis. Therefore, a positive test cannot differentiate between TB infection and active tuberculosis. A positive skin test does not indicate that the child is immune to tuberculosis (i.e. does not mean that the child is protected against tuberculosis).

A positive Mantoux skin test indicates tuberculous infection but not necessarily tuberculosis.

3-8 What are the causes of a false-positive Mantoux?

If the Mantoux test is read incorrectly it may be reported as positive when in fact it is negative. This may occur when the extent of erythema (redness) is measured instead of the extent of induration (swelling). Furthermore if the extent of induration is measured in the vertical plane it may be reported as positive when in fact it is negative. Therefore, measure the Mantoux reaction correctly i.e. the size of induration in the transverse diameter and report the result in millimetres.

Another reason for a false-positive Mantoux is infection with non-tuberculous mycobacteria. This is very uncommon in children, but may be considered if the child fails to respond to standard anti-tuberculous therapy.

3-9 What is the meaning of a negative skin test?

A negative skin test in a well-nourished, thriving child who is HIV negative usually means that the child has not been infected with TB bacilli and does not have tuberculosis.

However, the test may be falsely negative in spite of active infection with TB bacilli if the child’s immune system cannot react to the tuberculin in the following situations:

Therefore the Mantoux skin test may be negative even though the child has TB infection or tuberculosis (a false-negative test).

The Mantoux test may be negative in children with severe acute malnutrition, HIV infection, severe or disseminated tuberculosis or soon after measles.

3-10 How should PPD be stored?

To obtain an accurate Mantoux test the PPD must be correctly stored and the intradermal injection must be given correctly. Failure of either may result in a negative test in a child with tuberculosis. The PPD must be stored away from light and heat as these will damage it. Freezing also damages PPD.

Therefore, PPD should be stored in a refrigerator between 2 and 8 °C (not in the freezer compartment) and kept in a cool bag while being transported. Once the vial has been opened, it should be kept cool and used within six hours as it is a live vaccine.

3-11 Are there other tests to identify sensitivity to TB bacilli?

There are blood tests called interferon-gamma release assays, which can accurately identify infection with TB bacilli. The World Health Organization (WHO) has recommended that these tests do not replace tuberculin skin tests in low- and middle-income countries. For this reason these tests have not been introduced in the public sector in South Africa even though they are being used in the private sector. As with tuberculin skin tests, interferon-gamma release assays cannot distinguish between TB infection and tuberculosis. Another reason for not introducing these tests in South Africa is that they are expensive.

These tests detect interferon-gamma which is released by lymphocytes (that were previously sensitised to Mycobacterium tuberculosis) when stimulated in a test tube with Mycobacterium tuberculosis-specific proteins Interferon-gamma release assays can accurately distinguish between Mycobacterium tuberculosis and BCG, but cannot differentiate between tuberculous infection and tuberculosis.

Identifying TB DNA and bacilli in sputum

3-12 How useful are sputum tests in diagnosing tuberculosis?

Sputum testing is very important as they are a way of identifying TB bacilli. If TB bacilli are identified in the sputum the child has pulmonary tuberculosis.

3-13 Which are the commonly used sputum tests?

Previously, smear microscopy was used as the initial laboratory test to identify TB bacilli in the sputum. However Xpert MTB/RIF has a higher detection rate and is faster than smear microscopy. Therefore in 2013 WHO recommended that Xpert MTB/RIF instead of smear microscopy be used as the initial laboratory test for children and adults with suspected tuberculosis. When diagnosing tuberculosis, a positive Xpert MTB/RIF test is a more accurate method than demonstrating sensitisation to TB bacilli (Mantoux skin test).

The Xpert MTB/RIF test is the recommended method of diagnosing TB.

3-15 How is a specimen of sputum obtained for examination?

This is easy in an adult or older child who is asked to cough up a sample of sputum into a clean plastic container. It is essential that a sample of sputum and not saliva is obtained. Usually two samples of coughed up sputum on consecutive days are collected, at least one being collected early in the morning before brushing teeth or eating or drinking anything. The method is very successful if careful instruction is given on how to produce a sputum specimen. However, it is far more difficult in young children. Therefore other methods have to be used in younger children, especially children below six years of age who usually swallow their sputum:

3-16 How is a sample of gastric aspirate obtained?

If a sputum sample cannot be coughed up by a child it is advised to collect a sample of gastric fluid which contains swallowed sputum. The gastric aspirate is best collected early in the morning before the first feed, the child having been nil per mouth for at least six hours. A nasogastric tube is passed through the infant’s nose and pushed down into the stomach. Once a sample of gastric aspirate is obtained, the tube is removed. This is an uncomfortable procedure for the child.

As gastric fluid is highly acid, 4% sodium bicarbonate in an equal volume to the gastric aspirate should be added to the specimen to neutralise the stomach acid. Otherwise the TB bacilli will be killed before they can be cultured. A sample of gastric aspirate should be collected on two consecutive mornings.

Sodium hydroxide (preferably with N-acetyl-L-cysteine) can also be added to the sample of sputum to liquefy it and kill any contaminants. A mucolytic makes it easier to centrifuge the specimen, which concentrates any TB bacilli.

3-17 How can saline be used to improve the chance of collecting sputum?

The saline-induced method of obtaining sputum in young children is very useful when they are not able to cough up a sample themselves. This technique requires extra equipment and staff with special training. At present it is mostly used in tertiary-care hospitals but may in the future be used more widely.

It is important that a good sample of sputum is obtained to increase the chance that any TB bacilli will be detected.

In a child who has had nothing to eat or drink for three hours, nebulised hypertonic (3%) saline is breathed in to mobilise sputum. The saline is an irritant that is used to induce a cough.
A demonstration video of the paediatric sputum induction procedure produced by Medicines Sans Frontiers can be found on YouTube.

3-18 How rapidly should the sputum sample be sent to a laboratory for examination?

The sputum sample should be sent as quickly as possible to the nearest TB laboratory. It is best if the sputum is evaluated by Xpert MTB/RIF or smear microscopy within a few hours of obtaining the sample. The sputum container must be clearly marked and properly closed with the patient’s name and kept out of direct sunlight.

3-19 Do sputum samples always contain TB bacilli in children with pulmonary tuberculosis?

Young children usually cough up fewer TB bacilli and less TB DNA than older children, adolescents and adults do. This is because they do not have ‘adult-type’ tuberculosis.

Adults with tuberculosis usually have cavities in their lungs and cough up very large numbers of TB bacilli and TB DNA. Therefore they are very infectious and are said to have ‘open TB’. Sputum microscopy can identify the most infectious patients.

Children with only a few TB bacilli in their sputum are said to have ‘paucibacillary TB’.

3-20 Is there a danger to the staff if they collect sputum samples?

Yes. It is therefore important that staff members are protected against inhaling TB bacilli when coughed-up sputum samples are collected. The room should be well ventilated or a negative-pressure ventilated facility and the staff should wear N95 respirators. Children can also wear masks to prevent the spread of small droplets which may contain TB bacilli. This is particularly important if the tuberculosis is resistant to first-line antibiotics. Although children spread fewer TB bacilli, they can still infect staff collecting sputum. Staff should wash their hands after collecting a sputum sample.

Evaluation by samples by Xpert MTB-RIF

3-21 What is Xpert MTB/RIF?

Xpert MTB/RIF is a PCR-based test that simultaneously detects the presence of DNA of Mycobacterium tuberculosis complex and screens for rifampicin resistance. By detecting TB DNA Xpert MTB/RIF is used to diagnose tuberculosis. Detection of rifampicin resistance allows early initiation of appropriate drug-resistant TB treatment. Xpert MTB/RIF is the recommended initial laboratory test for confirming a clinical diagnosis of tuberculosis.

Mycobacterium tuberculosis complex includes several Mycobacterium species, notably Mycobacterium tuberculosis. However, other mycobacteria that uncommonly cause infection in children such as Mycobacterium bovis and Mycobacterium bovis BCG will be detected by this test. Thus the test is not specific for Mycobacterium tuberculosis infection.

3-22 Which samples may be evaluated using Xpert MTB/RIF?

Because pulmonary TB is the common form of tuberculosis, sputum is the most frequent sample type evaluated by Xpert MTB/RIF. However, other samples may be evaluated including fine needle aspirates obtained from lymph nodes, cerebrospinal fluid from children with suspected TB meningitis, and pus aspirates. Because of low yields, Xpert MTB/RIF is not routinely performed on clear (non-purulent) pleural, pericardial and peritoneal fluids, or on urine.

3-23 How is a sample processed using Xpert MTB/RIF?

The sputum (ideally a volume of 2 ml), CSF, fine needle aspirate or pus sample is inserted manually into a single-use Xpert MTB/RIF cartridge, which contains all the test reagents. The cartridge is inserted into the Xpert MTB/RIF instrument. The PCR analysis is completed automatically by the Xpert MTB/RIF instrument. The result can be obtained within 2 hours of receiving the sample. However, because of high workload in the routine microbiology laboratory there may be a delay of a few days before the result is obtained.

3-24 How reliable is Xpert MTB/RIF?

Although Xpert MTB/RIF is able to confirm more cases of TB than smear microscopy, its overall performance in children is poor, confirming tuberculosis in less than 20% of children with clinical, radiological and Mantoux evidence of tuberculosis.

3-25 How should you respond to a positive Xpert MTB/RIF result?

3-26 Does a negative Xpert MTB/RIF test exclude tuberculosis?

No. A negative Xpert MTB/RIF test does not exclude tuberculosis. Further testing may be indicated. Where tuberculosis is strongly suspected on clinical and/or radiological grounds a second sample should be sent for TB culture, line probe assay and drug susceptibility testing.

A negative Xpert MTB/RIF result does not rule out tuberculosis.

The above information is based on the 2013 South African national guidelines for managing tuberculosis in children. Individual provinces may differ in their approach to processing sputum samples.

Sputum smear examination

3-27 How useful is a sputum smear examination in diagnosing tuberculosis?

Examining a sputum smear under a microscope (sputum smear microscopy) can identify TB bacilli in a small percentage of children with tuberculosis. It is the oldest test used in identifying patients with tuberculosis. It has been largely replaced by the Xpert MTB/RIF test.

3-28 How is a sputum smear examined for TB bacilli?

In order to see TB bacilli, a smear of the sputum is made on a glass slide. The smear is stained and then examined under a microscope. Two methods are used to stain and examine a sputum smear.

  1. The traditional method is to stain the smear with Ziehl-Neelsen stain and then look for TB bacilli under a light microscope. With this method the TB bacilli are first stained and then washed with acid. Unlike other types of bacteria, TB bacilli do not lose the stain when the slide is washed with acid. As a result the TB bacilli are called ‘acid-fast bacilli’ or ‘AFBs’. A result can be obtained in 20 minutes.
  2. A new method uses auramine-O stain before the smear is examined with a microscope using a blue light. The blue light makes the TB bacilli easy to see (they glow or fluoresce). This is a quicker and more accurate method. A result can be obtained in ten minutes. It is the preferred method for performing smear microscopy in South Africa.

Patients with a positive smear (TB bacilli are seen) are called ‘smear-positive’ patients. They are far more infectious than patients with tuberculosis who are ‘smear negative’. The more TB bacilli that are seen, the more infectious the patient is to others.

Patients are called ‘smear positive’ if TB bacilli can be seen in their stained sputum.

With the auramine-O method a blue light-emitting diode (LED) can be used instead of an expensive mercury vapour lamp. The accuracy of microscopy depends on the number of TB bacilli, the expertise of the examiner and the time spent looking for TB bacilli.

3-29 How reliable is a sputum smear examination in diagnosing tuberculosis?

Unfortunately it is not very reliable in children as many children with pulmonary tuberculosis have few TB bacilli in their sputum. If the sputum smear examination is positive the child has pulmonary tuberculosis. However a negative smear does not rule out tuberculosis.

More than 90% of children with pulmonary tuberculosis will have a negative sputum smear because they have very few bacilli in their sputum.

3-30 How is smear microscopy used in South Africa?

Since the introduction of Xpert MTB/RIF, smear microscopy is only done in children who are Xpert MTB/RIF positive. Smear microscopy is done on a second sputum sample. If the smear microscopy is positive then further testing will be done during TB treatment and towards the end of TB treatment to document sputum conversion and cure. If the initial smear microscopy is negative, no further sputum samples are tested by smear microscopy during the course of anti-tuberculous treatment.

According to the 2013 South African national guidelines for managing paediatric tuberculosis, children who are smear-positive on sputum testing should be monitored by smear microscopy during TB treatment. One sputum sample should be collected for smear examination at the end of the intensive phase of TB treatment to evaluate smear conversion and another sputum sample should be collected at the end of the continuation phase of treatment to evaluate treatment outcome i.e. to establish whether or not the child was cured.

Culture for TB bacilli and drug susceptibility testing

3-31 Can TB bacilli be cultured from children with tuberculosis?

Yes, TB bacilli can be cultured (grown), especially from children with pulmonary tuberculosis. However it is not possible to culture TB bacilli from all children with tuberculosis.

3-32 How is sputum collected for sputum culture?

The same sputum collection methods that are used for sputum smear examination are also used for sputum culture. It is important to keep the sputum specimen cool at 4–10 °C and get it to the TB laboratory as soon as possible.

3-33 How is sputum cultured?

Either a solid or liquid culture medium is used. It may take four to eight weeks to get a positive culture on solid medium although the growth of TB bacilli is faster at two to three weeks with a liquid (broth) medium. This long wait is the main problem with TB culture. However, an advantage is that drug sensitivity or resistance testing can be done if a positive culture is obtained.

Like sputum smear staining, sputum culture in children is not as sensitive as in adults, as there are usually are far fewer TB bacilli in the sputum.

In South Africa, liquid culture systems are used for culturing Mycobacterium tuberculosis.

3-34 Is sputum culture more sensitive than Xpert MTB/RIF examination?

Yes. Culture is more accurate than both Xpert MTB/RIF examination and smear microscopy for confirming tuberculosis in children. The culture may be positive when Xpert MTB/RIF is negative in a child or adult with tuberculosis. Therefore TB culture is important, especially if the Xpert MTB/RIF result is negative in a child with a history, clinical examination, Mantoux skin test and chest X-ray suggesting tuberculosis.

3-35 Do all children with pulmonary tuberculosis have a positive sputum culture for TB bacilli?

Unfortunately, no. Due to the small number of TB bacilli in their sputum, many children with tuberculosis will have both negative Xpert MTB/RIF and culture results. Therefore the diagnosis of tuberculosis in children may have to be made on the history, clinical examination, chest X-ray and Mantoux test.

In more than 80% of children with tuberculosis, the diagnosis cannot be confirmed by sputum culture.

3-36 May Xpert MTB/RIF be positive and TB bacilli cultured from other sites?

Yes. Xpert MTB/RIF can be positive on CSF (cerebrospinal fluid), lymph node aspirates of lymph nodes, pus and in tissue biopsies. TB bacilli may also be cultured from these sources.

3-37 What is drug susceptibility testing?

Once Mycobacterium tuberculosis has been isolated from a patient sample, it is important to determine whether or not the isolate is susceptible to the drugs commonly used to treat tuberculosis. Xpert MTB/RIF is able to identify whether or not patients have tuberculosis that is resistant to rifampicin. However, further testing is required in patients to determine whether or not the TB bacilli are resistant to other TB drugs. Drug susceptibility testing is done in the microbiology laboratory using a combination of molecular and culture-based methods. Drug susceptibility testing will establish whether the patient has drug-susceptible or drug-resistant tuberculosis. If the patient has drug-resistant tuberculosis then drug susceptibility testing will also determine the type of drug-resistance. This is important so that the correct anti-tuberculous therapy may be prescribed.

Chest X-ray

3-38 How useful is a chest X-ray in diagnosing tuberculosis?

A chest X-ray is one of the most important special investigations in diagnosing tuberculosis, especially in children with a negative sputum test. In these children it may be the only way of confirming a diagnosis of pulmonary tuberculosis.

Chest X-ray is very important in diagnosing tuberculosis.

3-39 What chest X-ray views are needed?

It is important to get good quality frontal X-rays. A lateral view is useful but not essential.

3-40 What is the typical chest X-ray appearance in primary tuberculosis?

In pulmonary tuberculosis in children the most common finding on chest X-ray is hilar lymph node enlargement only.

3-41 What complications of primary tuberculosis can be seen on a chest X-ray?

3-42 Are the chest X-ray findings easy to interpret in children with tuberculosis?

While the X-ray findings may be typical of tuberculosis, the diagnosis may be difficult. Tuberculosis may present with a wide range of appearances while other chest conditions may look like tuberculosis. The chest X-ray is often particularly difficult to interpret in children with HIV infection. Unlike acute bacterial or viral pneumonia, the chest X-ray findings do not rapidly disappear when treatment is started. Diagnosis should not be made on chest X-ray alone.

Sometimes a short trial of treatment for acute pneumonia with a drug such as amoxicillin, which is not effective for tuberculosis, may be used if the clinical diagnosis is uncertain. A failure to improve both clinically and on chest X-ray would support a diagnosis of tuberculosis.

Fine needle aspiration of a lymph node

3-43 When is a fine needle aspiration indicated?

Enlarged lymph nodes (lymphadenopathy) or inflamed or fluctuant lymph nodes in the neck are common in childhood tuberculosis. Aspiration with a thin needle is a very useful way of detecting tuberculosis. Lymph node aspirates can be used to identify TB DNA by Xpert MTB/RIF and culture TB bacilli. The same tests can be done on pus from a discharging sinus.

Smear microscopy may also be used to identify TB bacilli in lymph node aspirates. However, this is no longer the preferred method in South Africa.

3-44 How is a needle aspiration performed?

Aspiration is done with a 22 or 23 gauge needle. Instead of an aspiration, sometimes the whole node may be excised and examined for TB bacilli and historical evidence (caseating granulomata) of tuberculosis. It is important not to incise (cut into) the node as this may lead to sinus formation.

Lumbar puncture

3-45 How is the clinical diagnosis of tuberculous meningitis confirmed?

If the symptoms and clinical signs suggest tuberculous meningitis, a lumbar puncture must be done to obtain a sample of cerebrospinal fluid (CSF) for examination. Patients with tuberculous meningitis have a raised CSF protein and low glucose concentration with many white cells, especially lymphocytes. Cerebrospinal fluid should also be tested by Xpert MTB/RIF for the presence of TB DNA, and may also be cultured for TB bacilli.

A lumbar puncture is usually contraindicated until after a computed tomography (CT) scan is done.

Additional investigations

3-46 Can advanced radiology be used to support the diagnosis of tuberculosis?

Yes, ultrasound, CT scan and MRI studies can assist with the diagnosis of extra-pulmonary tuberculosis. For example, if abdominal tuberculosis is suspected on clinical grounds, an ultrasound may be able to confirm hepatosplenomegaly, ascites and/or identify enlarged abdominal lymph nodes, features of abdominal tuberculosis. Similarly, if TB meningitis is suspected a CT scan of the head may identify features of TB meningitis such as hydrocephalus, basement meningeal enhancement and cerebral infarcts.

3-47 What other investigations may assist the diagnosis of tuberculosis?

If the child has abdominal ascites than an ascitic tap can be performed and biochemical analysis, microscopy and TB culture performed on the fluid. Similarly, if the child has a pericardial effusion then pericardial fluid may be obtained and analysed. Tissue obtained from a diseased organ at surgery or biopsy may also be evaluated histologically for evidence of tuberculosis and cultured for Mycobacterium tuberculosis.

Screening for HIV

3-48 Should all patients with suspected tuberculosis be screened for HIV?

Yes. This is important as many children with tuberculosis will also have HIV infection. Usually the HIV Rapid test can be used to screen children. However children under 18 months with a positive HIV Rapid test must also have a PCR test to determine whether they are HIV infected. A positive PCR result should be confirmed with a second PCR test. In children more than 18 months of age, a positive rapid test result should be confirmed with a second rapid test.

Tuberculosis usually spreads more rapidly if the child is also HIV infected. Therefore it is important that all children with tuberculosis be screened for HIV infection. There are three important reasons to screen for HIV.

  1. The diagnosis of tuberculosis is more difficult in children with HIV infection. The Mantoux test may be falsely negative while the clinical signs and chest X-ray appearance may be similar in tuberculosis and other HIV-related lung disease.
  2. All HIV-infected children treated for tuberculosis need to be referred for antiretroviral treatment.
  3. There are many drug interactions when children are treated for both tuberculosis and HIV. Therefore the medication, especially antiretroviral treatment, may require adjustment.

All patients with suspected tuberculosis must be screened for HIV infection.

Case study 1

A young child, whose mother has pulmonary tuberculosis, is brought to a clinic where the nurse performs a Mantoux skin test. BCG immunisation is recorded in the child’s Road-to-Health booklet. The mother is asked to bring the child back the following day so that the skin test can be read.

1. What is a Mantoux skin test?

Tuberculin (protein from dead TB bacilli) is injected into the skin. Later the site is examined for a reaction (an area of swelling).

2. When should the test be read?

48 to 72 hours (two to three days) after the test is done. It is too soon to read the test the following day as the skin reaction may not be fully developed yet.

3. How is the test read?

Using a ruler, the largest transverse (across the arm) diameter of the induration is read in millimetres.

4. What is a positive test?

In immunosuppressed children (including HIV-infected children and children with severe acute malnutrition) the Mantoux is positive if the transverse diameter of induration is greater than 5 mm. In all other children, whether or not they received BCG vaccine, the Mantoux is regarded as positive if the transverse diameter of induration is 10 mm or more.

5. What does this result mean?

The child has TB infection. However a positive test cannot indicate whether the child also has tuberculosis.

6. Could the result be due to BCG immunisation?

No, as BCG may produce some induration but not a positive reaction.

Case study 2

A 4 year old child with suspected tuberculosis is referred to a TB clinic for sputum examination. It is also suggested that the child has a chest X-ray.

1. How should a sputum sample be obtained?

Unlike adults and older children, young children usually cannot cough up a sample of sputum. Therefore it has to be obtained by gastric aspirate or saline-induced collection.

2. What tests will be done on the sputum?

An Xpert MTB/RIF test and culture. If the Xpert MTB/RIF is positive smear microscopy is done to determine whether visible TB bacilli are present. In children who have TB bacilli on sputum examination (positive smear microscopy), further smear microscopy should be done to determine whether sputum conversion has occurred at the end of the intensive phase of TB treatment and to evaluate outcome at the end of TB treatment i.e. to establish whether or not the child was cured.

3. Does a negative Xpert MTB/RIF result exclude pulmonary tuberculosis?

No, as a negative Xpert MTB/RIF result is common in children with tuberculosis. However a positive smear will confirm a clinical diagnosis of tuberculosis.

4. Is it worth asking for a culture if the Xpert MTB/RIF result is negative?

Yes, as the culture is more sensitive. In children the culture may be positive when the Xpert MTB/RIF result is negative. This is because there often are only a few TB bacilli in the sputum of children.

5. How long does it take to get a culture result?

Up to four weeks with liquid culture medium.

6. Is a chest X-ray also necessary?

Not if the Xpert MTB/RIF result is positive in older children. If the Xpert MTB/RIF result is negative a chest X-ray is useful to confirm the diagnosis. The chest X-ray can be difficult to interpret in children with tuberculosis, especially if they are also infected with HIV.

Case study 3

A young child with enlarged cervical nodes presents at a district hospital. The medical officer phones the referral hospital for advice on what investigations to do. There is a strong family history of tuberculosis.

1. What samples are required?

Obtain a sputum sample and a fine needle aspirate from an enlarged lymph node. The sputum sample will probably be collected by gastric aspirate or after sputum induction with nebulized hypertonic saline.

2. What investigations are needed?

Xpert MTB/RIF result examination and culture on both samples.

3. How should the gastric aspirate be prepared?

An equal volume of 4% sodium bicarbonate should be added to the gastric aspirate to neutralise the stomach acid. This will increase the chance of a positive culture. The sample should be kept cool and sent to the laboratory as soon as possible.

4. Is it not easier to simply take a chest X-ray?

In children it is difficult to diagnose tuberculosis on a chest X-ray alone. It is important to take a history, assess growth, perform a Mantoux test, and screen for tuberculosis on sputum or some other sample such as a lymph node aspirate, pleural aspirate or cerebrospinal fluid using Xpert MTB/RIF and culture. Many children with peripheral (cervical) lymphadenopathy have a normal chest X-ray.

5. What stains are used on a sputum smear?

Either a Ziehl-Neelsen or auramine-O stain.

6. Is it dangerous for health workers to collect a sputum sample?

It is possible to become infected with TB bacilli while collecting a coughed-up sputum sample. Therefore the health worker should wear a N95 respirator during the procedure, choose a well-ventilated or negative-pressure ventilated space and wash their hands well afterwards. This is particularly important in communities where drug-resistant tuberculosis is common.

The five most important ‘take-home’ messages

  1. A positive Mantoux skin test indicates TB infection and not necessarily tuberculosis.
  2. A diagnosis of tuberculosis in a child depends on a history of exposure to tuberculosis, chronic symptoms of cough, a positive Mantoux skin test and typical chest X-ray findings. Not all have to be present to diagnose tuberculosis.
  3. Tuberculosis is confirmed by the Xpert MTB/RIF test or by culturing TB bacilli in the sputum, gastric aspirate or other sample.
  4. A chest X-ray is important for diagnosing tuberculosis but not monitoring the response to treatment.
  5. HIV screening is essential in all children suspected of having tuberculosis.