Approach

The diagnosis may be obvious in some cases but is frequently difficult. A high level of suspicion is important in evaluating a patient with risk factors. Diagnosis confirmation requires culturing of Mycobacterium tuberculosis. Delays in diagnosis and initiation of therapy are associated with transmission of disease and increased mortality.[12]

If suspicion for disease is high, the patient should be isolated (at home or in a negative-pressure room in a hospital) until 5 days to 2 weeks of therapy has been completed. Active TB, confirmed or highly suspected, is a reportable condition to the local health authorities.

Clinical history and risk factors

The possibility of TB should be considered in any person with risk factors for TB exposure, who has suggestive symptoms (e.g., fever, malaise, pleuritic chest pain, cough longer than 2-3 weeks, night sweats, and weight loss, hemoptysis, psychological symptoms, clubbing, erythema nodosum) or chest x-ray abnormalities. Although the presence of upper lobe infiltrates is characteristic of the disease, atypical chest x-ray presentation is common among children, and among people who are immunocompromised, have HIV infection, or have diabetes.[Figure caption and citation for the preceding image starts]: Pulmonary TB with cavitationFrom the personal collection of David Horne and Masahiro Narita; used with permission [Citation ends].com.bmj.content.model.Caption@4a1e5d48

Diagnostic investigations

Investigations for active infection include chest x-ray, three sputum samples obtained for acid-fast bacilli (AFB), nucleic acid amplification testing (NAAT), full blood count, and electrolytes (e.g., sodium). If the patient is unable to spontaneously produce sputum, it should be induced (with appropriate precautions to prevent transmission) or obtained via bronchoscopy or gastric aspirate.[12] Stained smears should be made from sputum specimens to identify AFB, as this is the first bacteriological evidence of infection and gives an estimate of how infectious the patient is. If AFBs are seen on smear, therapy should be started and the patient maintained in isolation.

Sputum culture supports the diagnosis of TB, is more sensitive and specific than smear staining, facilitates identification of the mycobacterium species by nucleic acid hybridisation or amplification, and evaluates drug sensitivity. Broth culture systems allow for rapid growth and detection in 1 to 3 weeks as opposed to 4 to 8 weeks by a solid medium.[12]

NAAT should be performed on at least one respiratory specimen when a diagnosis of TB is being considered. NAAT may speed the diagnosis in smear-negative cases and may be helpful to differentiate non-tuberculous mycobacteria when sputum is AFB smear positive but NAAT negative.[33] Genotyping might be considered useful in outbreaks of TB to identify recent transmission of TB, especially when contact had not been identified in the course of epidemiological investigations. Several rapid NAATs are available and some are also able to detect genes encoding resistance to TB drugs.[30][34]​​​​​​​[35][36][37][38][39] A positive NAAT is adequate for initiation of antituberculosis treatment. When the patient has previously been treated for TB, especially within the prior 2 years, a positive NAAT may represent a false-positive result.

Use of stool samples is an alternative to respiratory specimens in diagnosis of pulmonary TB (sputum is swallowed and M tuberculosis may pass through the gastrointestinal tract). One systematic review evaluating AFB-smear, culture, and NAAT (polymerase chain reaction [PCR]) testing of stool in pulmonary TB found a pooled sensitivity of one or more of the three tests was 79.1% (95% CI 61.5 to 92.5).[40] The sensitivity of stool microscopy, PCR, and culture was 41.1% (95% CI 24.9 to 58.2), 89.7% (95% CI 81.4 to 95.9), and 38.0% (95% CI 26.2 to 50.6), respectively.[40]

The World Health Organization (WHO) recommends that in children with signs and symptoms of pulmonary TB, the NAAT Xpert Ultra should be used for initial diagnostic testing and detection of rifampicin resistance on sputum, nasopharyngeal aspirate, gastric aspirate, or stool, rather than smear microscopy/culture and phenotypic drug susceptibility testing (DST).[30]​​​​

CT of the chest, although not done routinely, may be of use to exclude other pathology: for example, cancer.

Lateral flow tests that detect lipoarabinomannan (LAM) antigen in urine have emerged as potential point-of-care tests. One Cochrane review found the lateral flow urine lipoarabinomannan (LF-LAM) assay to have a sensitivity of 42% in diagnosing TB in HIV-positive individuals with TB symptoms, and 35% in HIV-positive individuals not assessed for TB symptoms.[41] [ Cochrane Clinical Answers logo ] [ Cochrane Clinical Answers logo ] WHO recommends that LF-LAM can be used to assist in the diagnosis of active TB in HIV-positive adults, adolescents, and children.[30][34]​​​​​ This approach is supported by another Cochrane review, which found reductions in mortality and an increase in treatment initiation with use of LF-LAM in inpatient and outpatient settings.[42] Culture would still be required for drug susceptibility testing (DST).

It is recommended that all patients who have TB should be tested for HIV within 2 months of diagnosis.

Negative smear or cultures

Around 40% to 50% of cases are AFB smear negative and 15% to 20% have negative cultures. In patients where there is a strong clinical suspicion of TB, especially if the tuberculin skin test is positive (reaction of ≥5 mm induration), empirical TB therapy may be tried before laboratory confirmation of infection. Clinical and radiographic improvement with appropriate antituberculosis treatment supports the diagnosis. NAAT of sputum cultures may also be helpful in this situation. Bronchoscopy can be done to obtain bronchoalveolar lavage samples or transbronchial biopsy, and gastric aspirate can be used in patients unable to provide an adequate sputum sample, such as young children. In patients where suspicion for active TB is low and smears are negative for AFB, it is acceptable to wait for the results of AFB culture or repeat chest x-ray before starting treatment.[9]

Patients with smear-negative disease may be infectious, although the risk of transmission is lower than in smear-positive disease.[43] If the suspicion of TB is high, consideration should be given to starting antituberculous medicines prior to laboratory confirmation.

Susceptibility testing

In adequately resourced settings, culture-based DST is routinely performed on initial M tuberculosis isolates. A limitation of culture-based DST is that it can take >2 weeks to grow the isolate for testing. When there is a higher suspicion for drug resistance, then rapid molecular DST may be appropriate to guide treatment. Rapid molecular DST for rifampicin with or without isoniazid using the respiratory specimens of persons who are either AFB smear positive or NAAT positive should be considered in patients who:

  • Have been treated for TB in the past, or

  • Were born in or have lived for at least 1 year in a foreign country with at least a moderate TB incidence (≥20 per 100,000) or a high primary multidrug-resistant (MDR) TB prevalence (≥2%), or

  • Are contacts of patients with MDR TB, or

  • Are HIV-positive.

When Xpert MTB/RIF or Xpert Ultra (rapid molecular tests endorsed by the World Health Organization) are used as part of TB diagnosis, rifampicin resistance will be automatically assessed.[12][34]​​​ Cochrane reviews of Xpert MTB/RIF and Xpert Ultra found that they provide accurate results for rifampicin‐resistant and MDR TB.[35][36][39] Xpert MTB/XDR assesses resistance to isoniazid, fluoroquinolones, injectables (amikacin, kanamycin, capreomycin), and ethionamide. One Cochrane review found Xpert MTB/XDR is accurate for detecting isoniazid and fluoroquinolone resistance.[37]

Line probe assays (LPAs) are strip-based tests that can detect TB and determine drug resistance profiles. LPAs are recommended by WHO only for detecting resistance to anti-TB drugs.[34]​​ WHO also now includes next-generation sequencing (NGS) in its recommendations for DST in confirmed pulmonary TB disease.[34]​​ NGS is a new class of test that can be used for the detection of resistance to a broad list of drugs.​​

Tuberculin skin test and interferon gamma release assays

Investigations for latent infection in a person exposed to M tuberculosis but without signs of active TB are based on the tuberculin skin test (TST) or interferon gamma release assays (IGRAs).The TST and IGRA measure the response of T-cells to TB antigens. As false-negative results occur in 20% to 25% of patients with active pulmonary TB, these tests should not be used alone to exclude a diagnosis of active TB.[44] The American Academy of Pediatrics advises that a negative result of either TST or IGRA should be considered especially unreliable in a child younger than 3 months.[45] Interpretation of the TST depends on patient characteristics including immunocompetence and vaccination status. For patients with normal immunity and no additional risk factors, induration of ≥15 mm in diameter is taken to mean a positive result, but a smaller diameter is used as a cut-off in people with additional risk factors and in children.[12][30]​​​​​​​​​[46]​ An IGRA may be used in place of a TST in all situations in which TSTs are used to diagnose latent infection, though TST may be preferred in children aged younger than 2 years.[45][47]​ An IGRA is preferred in individuals with a history of bacille Calmette-Guérin vaccination due to superior specificity.[46] In addition, IGRA is preferred for testing persons from groups that historically have low rates of returning to have TSTs read.

Targeted testing for latent TB infection is recommended by the US Centers for Disease Control and Prevention/American Thoracic Society as part of strategic control and reduction of TB. High-risk groups include people with HIV, intravenous drug users, healthcare workers who serve high-risk populations, and contacts of individuals with pulmonary TB.[12] Testing for latent TB infection should also be performed in patients prior to tumour necrosis factor-alpha antagonist therapy.[48]

TB antigen-based skin tests (TBSTs) are a new class of tests that have been developed to measure the cell-mediated immunological response to M tuberculosis specific antigens. The WHO recommends that TBSTs may be used to test for latent TB infection, reporting that the diagnostic accuracy of TBSTs is similar to that of IGRAs and greater than that of the TST.[49]

Use of this content is subject to our disclaimer