Tuberculosis (TB) Workup

Updated: Nov 09, 2017
  • Author: Thomas E Herchline, MD; Chief Editor: Michael Stuart Bronze, MD  more...
  • Print

Approach Considerations

The primary screening method for tuberculosis (TB) infection (active or latent) is the Mantoux tuberculin skin test with purified protein derivative (PPD). An in vitro blood test based on interferon-gamma release assay (IGRA) with antigens specific for Mycobacterium tuberculosis can also be used to screen for latent TB infection. IGRA assays offer certain advantages over tuberculin skin testing. [44, 45]

Obtain the following laboratory tests for patients with suspected TB:

  • Acid-fast bacilli (AFB) smear and culture - Using sputum obtained from the patient
  • HIV serology in all patients with TB and unknown HIV status

AFB stain is quick but requires a very high organism load for positivity, as well as the expertise to read the stained sample. This test is more useful in patients with pulmonary disease. Other diagnostic testing may need to be considered, as a delay in diagnosis can increase patient mortality. Traditional mycobacterial cultures require weeks for growth and identification. Newer technologies allow identification within 24 hours.

Obtain a chest radiograph to evaluate for possible associated pulmonary findings. If chest radiography findings suggest TB and a sputum smear is positive for AFB, initiate treatment for TB. A computed tomography (CT) scan of the chest may help to better define abnormalities in patients with vague findings on chest radiography. See the images below.

Axial noncontrast enhanced computed tomography wit Axial noncontrast enhanced computed tomography with pulmonary window shows a cavity with an irregular wall in the right apex of a 37-year-old man who presented with cough and fever (same patient as above).
Coronal reconstructed computed tomography image sh Coronal reconstructed computed tomography image shows the right apical cavity in a 37-year-old man who presented with cough and fever (same patient as above).
Axial chest computed tomography without intravenou Axial chest computed tomography without intravenous contrast with pulmonary window setting shows a right apical thick-walled cavity and surrounding lung consolidation in a 43-year-old man who presented with cough and fever (same patient as above).
Coronal reconstructed computed tomography image sh Coronal reconstructed computed tomography image shows the consolidated, partially collapsed right upper lobe with a cavity that is directly connected to a bronchus in a 43-year-old man who presented with cough and fever (same patient as above).
Axial chest computed tomography without intravenou Axial chest computed tomography without intravenous contrast with pulmonary window setting through the mid-chest shows a large, irregular-walled cavity with nodules and air-fluid level and two smaller cavities in a 43-year-old man who presented with cough and hemoptysis (same patient as above). Small, patchy peripheral opacities are also present in the left lower lobe. In the right mid-lung, nodular opacities are in a tree-in-bud distribution, suggestive of endobronchial spread.
Coronal reconstructed computed tomography image sh Coronal reconstructed computed tomography image shows the lingular cavity with irregular nodules and right mid-lung nodular opacities in a 43-year-old man who presented with cough and hemoptysis (same patient as above).

Technetium-99m (99m Tc) methoxy isobutyl isonitrile single-photon emission CT (SPECT) scanning for solitary pulmonary nodules yields a high predictive value for distinguishing TB from malignancy. Therefore, it has the potential to serve as a low-cost alternative when positron emission tomography (PET) scanning is not available, especially in endemic areas. [46]

Symptoms and radiographic findings do not differentiate multidrug-resistant TB (MDR-TB) from fully susceptible TB. Suspect MDR-TB if the patient has a history of previous treatment for TB, was born in or lived in a country with a high prevalence of MDR-TB, has a known exposure to an MDR-TB case, or is clinically progressing despite standard TB therapy.

Extrapulmonary TB

Extrapulmonary involvement occurs in one fifth of all TB cases, although 60% of patients with extrapulmonary manifestations of TB have no evidence of pulmonary infection on chest radiograph or sputum culture. Biopsy of bone marrow, liver, or blood cultures is occasionally necessary and may be helpful. Ocular TB can be especially difficult to identify, owing to its mimicry of other disorders and its lack of accessible sampling; a high index of suspicion is required.

The hallmark of extrapulmonary TB histopathology is the caseating granuloma, consisting of giant cells with central caseating necrosis. Rarely, if ever, are any TB bacilli seen.

Altered mental status, neck stiffness, decreased level of consciousness, increased intracranial pressure, and cranial nerve involvement can indicate tuberculous meningitis or tuberculoma. If these conditions are suspected, performing a lumbar puncture for evaluation of the cerebrospinal fluid is necessary. In addition, a tuberculoma can be substantiated on the basis of an increase in intracranial pressure and findings on CT or magnetic resonance image (MRI) scans.

If vertebral (Pott disease) or brain involvement is suspected, it is important to consider that a delay in treatment could have severe repercussions for the patient (ie, compression of the spinal cord and/or paraplegia). Consequently, further evaluation is necessary with CT scanning or MRI.

Urinalysis and urine cultures can be obtained for patients with genitourinary complaints. Although patients are often asymptomatic, significant pyuria and/or hematuria with no routine bacterial organisms mean that a urine culture for acid-fast bacilli should be obtained.


Pregnancy provides an opportunity to screen for TB; all pregnant women can undergo tuberculin skin testing. If the results are positive, chest radiography can be performed with lead shielding. Chest radiography should not be delayed during the first 3 months of pregnancy in patients with suggestive symptoms.

TB in children

For congenital TB, the best diagnostic tests are the pathologic and histologic examination of the placenta, as well as a placental culture. Mycobacterial blood cultures of the newborn may also be helpful. Treatment may be necessary until placental culture results are negative.

Postnatal TB in infants is contracted via the airborne route. The most common findings of postnatal TB include adenopathy and a lung infiltrate. However, the chest radiographic findings may be normal in infants with disseminated disease.

Chest radiographs in children with TB may show only hilar lymphadenopathy or a patchy infiltrate. Gastric aspirates or biopsies are not necessary if positive cultures have been obtained from the source case. Go to Pediatric Tuberculosis for complete information on this topic.

Patients with HIV infection

Individuals infected with HIV are at increased risk for TB, beginning within the first year of HIV infection. [47] All patients who are diagnosed with active TB and who are not known to be HIV positive should be considered for HIV testing.

The initiation of antiretroviral therapy (ART) decreases the risk of developing TB in these patients, [48] although the TB risk remains higher in the first 3 months of ART. The highest risk is in patients with the following factors [49] :

  • Baseline CD4+ count of less than 200/μL
  • Higher baseline HIV-1 ribonucleic acid (RNA) level - Relative hazard 1.93 for every log increase in baseline HIV-1 RNA
  • History of injection drug use
  • Male sex

In a study from Durban, South Africa, nearly 20% of patients starting ART had undiagnosed, culture-positive pulmonary TB. Neither cough nor acid-fast bacillus smear were sufficiently sensitive for screening. TB sputum cultures should be attempted before ART initiation in areas with a high prevalence of TB. [50]

Patients with TB must be tested for HIV, and patients with HIV need periodic evaluation for TB with tuberculin skin testing and/or chest radiography. Patients with HIV and a positive tuberculin skin test result develop active TB at a rate of 3-16% per year.

Patients with TB and HIV are more likely to have disseminated disease and less likely to have upper lobe infiltrates or classic cavitary pulmonary disease. Patients with a CD4 count of less than 200/μL may have mediastinal adenopathy with infiltrates.


Sputum Smear

Patients suspected of having TB should submit sputum for AFB smear and culture. Sputum should be collected in the early morning on 3 consecutive days. In hospitalized patients, sputum may be collected every 8 hours. [51] Early-morning gastric aspirate may also produce a good specimen, especially in children.

In patients without spontaneous sputum production, sputum induction with hypertonic saline should be attempted. [52] Fiberoptic bronchoscopy with bronchoalveolar lavage (and perhaps transbronchial biopsy) can be used if other attempts at obtaining sputum specimens are unsuccessful.

Ziehl-Neelsen staining of sputum is a simple 5-step process that takes approximately 10 minutes to accomplish. While highly specific for mycobacteria, however, this stain is relatively insensitive, and detection requires at least 10,000 bacilli per mL; most clinical laboratories currently use a more sensitive auramine-rhodamine fluorescent stain (auramine O).

TB detection following negative smear

The absence of a positive smear result does not exclude active TB infection. Approximately 35% of culture-positive specimens are associated with a negative smear result.

Jafari et al found that an M tuberculosis –specific enzyme-linked immunospot (ELISpot) assay can be used to differentiate TB cases with negative sputum smears from latent TB infection. In a prospective study of 347 patients suspected of having active TB who were unable to produce sputum or who had AFB-negative sputum smears, ELISpot testing of bronchoalveolar lavage fluid displayed a sensitivity and specificity of 91% and 80%, respectively, for the diagnosis of active pulmonary TB. [53]


Nucleic Acid Amplification Tests

Deoxyribonucleic acid (DNA) probes specific for mycobacterial ribosomal RNA identify species of clinically significant isolates after recovery. In tissue, polymerase chain reaction (PCR) amplification techniques can be used to detect M tuberculosis -specific DNA sequences and thus, small numbers of mycobacteria in clinical specimens. [54, 55]

Ribosomal RNA probes and DNA PCR assays allow identification within 24 hours. The DNA probes are approved for direct testing on smear-positive or smear-negative sputa. However, smear-positive specimens yielded higher sensitivity.

The CDC recommends performing one of these nucleic acid amplification tests when the diagnosis of pulmonary TB is being considered but has not yet been established, and when the test result would alter case management or TB control activities, such as contact investigations. The CDC recommends performing the test on at least 1 respiratory specimen. [56]

A retrospective cohort analysis of 2140 patients with suspected pulmonary TB found that Mycobacterium tuberculosis direct (MTD) nucleic acid amplification testing (NAAT) yielded improved diagnostic accuracy, shortened time to diagnosis, and reduced unnecessary treatment. In all study subpopulations examined (HIV-infected, homeless, substance abuser, and foreign-born), MTD had higher positive predictive value, sensitivity, and negative predictive value than no MTD, and in all subpopulations except homeless patients, MTD had higher specificity. In HIV-infected or homeless patients, MTD substantially reduced the cost of diagnosing or excluding TB, and in substance abusers, it cut the cost of excluding TB in those with smear-negative specimens. [57, 58]



Culture for AFB is the most specific test for TB and allows direct identification and determination of susceptibility of the causative organism. Access to the organisms, however, may require lymph node/sputum analysis, bronchoalveolar lavage, or aspirate of cavity fluid or bone marrow. In addition, obtaining the test results is slow (3-8 wk), and they have a very low positivity in some forms of disease.

Routine culture uses a nonselective egg medium (Lowenstein-Jensen or Middlebrook 7H10) and often requires more than 3-4 weeks because of the 22-hour doubling time of M tuberculosis. Radiometric broth culture (BACTEC radiometric system) of clinical specimens was found to significantly reduce the time (10-14 days) for mycobacterial recovery.

Newer broth culture media and systems for isolation, based on a fluorescent rather than a radioactive indicator, are available for use in clinical laboratories. The indicator is inhibited by oxygen; as mycobacteria metabolize substrates in the tubes and use the oxygen, the tube begins to fluoresce. [59]

Blood cultures

Blood cultures using mycobacteria-specific, radioisotope-labeled systems help to establish the diagnosis of active TB. However, mycobacterial bacteremia (bacillemia) is detectable using blood cultures only if specialized systems are used; these bacilli have specific nutrient growth requirements not met by routine culture systems.

Such blood cultures should be used for all patients with HIV infection who are suspected of having TB, because bacillemia is particularly prevalent in this population. If available, in fact, these cultures should be used for any patient highly suspected of having active TB.


Drug Susceptibility Testing

Positive cultures should be followed by drug susceptibility testing. Symptoms and radiographic findings do not differentiate MDR-TB from fully susceptible TB. Suspect MDR-TB if the patient has a history of previous treatment for TB, was born in or lived in a country with a high prevalence of MDR-TB, has a known exposure to an MDR-TB case, or is clinically progressing despite standard TB therapy. Susceptibilities should be repeated if cultures remain positive after 2 months, even when initial susceptibilities have not revealed any resistance.

DNA sequencing

Because conventional drug susceptibility tests for M tuberculosis take at least 3-8 weeks, Choi et al recommend direct DNA sequencing analysis as a rapid and useful method for detecting drug-resistant TB. In their clinical study, turnaround time of the direct DNA sequencing analysis was 3.8 +/- 1.8 days. The sensitivity and specificity of the assay were 63.6% and 94.6% for isoniazid, 96.2 and 93.9% for rifampin, 69.2% and 97.5% for ethambutol, and 100% and 92.6% for pyrazinamide, respectively. [60]

Automated molecular testing

An automated molecular test that uses sputum samples for the detection of M tuberculosis and resistance to rifampin has been developed. In studies conducted in low-income countries, the sensitivity for TB was 98.3% using a single smear-positive sputum sample. Sensitivity with a single smear-negative sputum sample was 76.9%, but it increased to 90.2% when 3 samples were tested. The test correctly identified 94.4% of rifampin-resistant organisms and 98.3% of rifampin-sensitive organisms. [61, 62]

MODS and TLA assays

Microscopic-observation drug susceptibility (MODS) and thin-layer agar (TLA) assays are inexpensive, rapid alternatives to conventional and molecular methods of TB drug susceptibility testing. The WHO has endorsed the MODS assay as a direct or an indirect test for rapid screening of patients with suspected MDR-TB. The evidence is insufficient to recommend the use of the TLA assay for rapid screening, but this assay is a promising diagnostic technique. [63]

Additional rapid tests

Other rapid tests are also available, such as BACTEC-460 (Becton-Dickinson), ligase chain reaction, and luciferase reporter assays (within 48 h) (Franklin Lakes). These tests have been developed for rapid drug-susceptibility testing, and results from these can be available within 10 days. Drug-resistance tests such as the FASTPlaque TB-RIF for rifampin resistance can be used after growth in semiautomated liquid cultures such as BACTEC-460; rifampin resistance can be used as a surrogate marker for isoniazid resistance.


Chest Radiography

Obtain a chest radiograph to evaluate for possible TB-associated pulmonary findings (demonstrated in the images below). A traditional lateral and posteroanterior (PA) view should be ordered. In addition, an apical lordotic view may permit better visualization of the apices and increase the sensitivity of chest radiography for indolent or dormant disease.

This radiograph shows a patient with typical radio This radiograph shows a patient with typical radiographic findings of tuberculosis.
Anteroposterior chest radiograph of a young patien Anteroposterior chest radiograph of a young patient who presented to the emergency department (ED) with cough and malaise. The radiograph shows a classic posterior segment right upper lobe density consistent with active tuberculosis. This woman was admitted to isolation and started empirically on a 4-drug regimen in the ED. Tuberculosis was confirmed on sputum testing. Image courtesy of Remote Medicine (
Lateral chest radiograph of a patient with posteri Lateral chest radiograph of a patient with posterior segment right upper lobe density consistent with active tuberculosis. Image courtesy of Remote Medicine (
This chest radiograph shows asymmetry in the first This chest radiograph shows asymmetry in the first costochondral junctions of a 37-year-old man who presented with cough and fever. Further clarification with computed tomography is needed.
This posteroanterior chest radiograph shows right This posteroanterior chest radiograph shows right upper lobe consolidation with minimal volume loss (elevated horizontal fissure) and a cavity in a 43-year-old man who presented with cough and fever.
The posteroanterior chest radiograph shows a large The posteroanterior chest radiograph shows a large cavity with surrounding consolidation in the lingular portion of the left upper lobe in a 43-year-old man who presented with cough and hemoptysis. There are also a few nodular opacities in the right mid-lung zone.

The chest film is also useful to screen for sarcoidosis, which closely imitates the clinical course of ocular TB. Radiologists look more decisively for signs of TB or sarcoid if the requesting physician specifies an interest in these.

Chest radiographs may show a patchy or nodular infiltrate. TB may be found in any part of the lung, but upper lobe involvement is most common. The lordotic view may better demonstrate apical abnormalities.

The following patterns may be seen on chest radiographs:

  • Cavity formation - Indicates advanced infection and is associated with a high bacterial load
  • Noncalcified round infiltrates - May be confused with lung carcinoma
  • Homogeneously calcified nodules (usually 5-20 mm) - Tuberculomas; represent old infection rather than active disease
  • Miliary TB - Characterized by the appearance of numerous small, nodular lesions that resemble millet seeds on chest radiography (go to Miliary Tuberculosis for complete information on this topic)

Chest radiography consistent with TB indicates active disease in the symptomatic patient even in the absence of a diagnostic sputum smear. Similarly, normal chest radiographic findings in the symptomatic patient do not exclude TB, particularly in a patient who is immunosuppressed.

Primary TB

In primary active TB, radiographic features of pulmonary tuberculosis are nonspecific, sometimes even normal. The chest radiograph typically shows a pneumonialike picture of an infiltrative process in the middle or lower lung regions, often associated with hilar adenopathy and/or atelectasis.

Primary TB is more likely to mimic the appearance of routine community-acquired pneumonia (CAP) on chest radiography than is reactivation TB. Studies have shown that primary TB and CAP may be associated with pleural effusion and cavitation.

Reactivation TB

In classic reactivation TB, pulmonary lesions are located in the posterior segment of the right upper lobe, the apicoposterior segment of the left upper lobe, and the apical segments of the lower lobes. Cavitation is most common; healing of tubercular regions results in the development of a scar, with loss of lung parenchymal volume and calcification.

TB and HIV infection

In patients with HIV infection or another immunosuppressive disease, lesions are often atypical. Up to 20% of HIV-positive patients with active TB have normal chest radiographic findings.

Healed and latent TB

Old, healed TB presents a different radiographic appearance, with dense pulmonary nodules, with or without calcifications, in the hilar or upper lobes. Smaller nodules, with or without fibrotic scars, can be seen in the upper lobes. Nodules and fibrotic lesions are well demarcated, have sharp margins, and are dense.

Patients with nodular or fibrotic scars on chest radiography and positive PPD results should be treated as latent carriers. Calcified nodular lesions (granulomas) or apical pleural thickening indicate a lower risk of conversion.

Miliary TB

In disseminated/miliary TB, the chest radiograph commonly shows a miliary pattern, with 2-mm nodules that, histologically, are granulomas disseminated like millet seeds throughout the lung. However, chest radiographic patterns can vary and can include upper lobe infiltrates with or without cavitation.

Pleural TB

In pleural tuberculosis, the pleural space can be involved in 2 ways: (1) a hypersensitivity response can cause pleuritic pain and fever, or (2) an empyema can be present that can be seen on chest radiographs, with associated pleural effusions.

See the following articles for more information on TB imaging studies:



Tuberculin skin test

The primary screening method for TB infection (active or latent) is the Mantoux tuberculin skin test with PPD. Tuberculin sensitivity develops 2-10 weeks after infection and usually is lifelong. Tuberculin skin testing is based on the fact that TB infection induces a strong, cell-mediated immune response that results in a measurable delayed-type hypersensitivity response to intradermal inoculation of tuberculin PPD.

The PPD test involves an intradermal injection of 5 units of PPD (0.1 mL), preferably with a 26-, 27-, or 30-gauge needle. The results should be read between 48 and 72 hours after administration. In immunologically intact individuals, induration of less than 5 mm constitutes a negative result.

Population-based criteria for PPD positivity are as follows:

  • Cutoff of 5 mm or more induration - Patients who are HIV positive, have abnormal chest radiographic findings, have significant immunosuppression, or have had recent contact with persons with active TB
  • Cutoff of 10 mm or more induration - Patients who are intravenous drug users, residents of nursing homes, prisoners, impoverished persons, or members of minority groups
  • Cutoff of 15 mm or more induration - Patients who are young and in good health

Reactions in patients who have received the bacillus Calmette-Guérin (BCG) vaccine should be interpreted in the same way as the reactions above, regardless of BCG history, according to guidelines from the CDC. [64]

Interferon-gamma release assay

An in vitro blood test based on IGRA with antigens specific for M tuberculosis can also be used to screen for latent TB infection and offers certain advantages over tuberculin skin testing. [44, 45] The 2 currently available tests are (1) QuantiFERON-TB Gold In-Tube (QFT-GIT), an enzyme-linked immunosorbent assay (ELISA) based on ESAT-6, CFP-10, and TB 7.7 antigens, and (2) T-SPOT.TB, an enzyme-linked immunosorbent spot (ELISpot) assay based on ESAT-6 and CFP-10 antigens.

IGRA tests measure T-cell interferon-gamma response to antigens that are highly specific for M tuberculosis and absent from the BCG vaccine and M avium. [65] Overall, sensitivity and specificity of IGRA are comparable to those of tuberculin skin testing; however, unlike tuberculin skin testing, these tests do not require a second visit for a reading. Results are reported as positive, negative, or indeterminate. Patients with an indeterminate result may have evidence of immunosuppression and may be nonreactive on skin testing. [66]

Advantages of IGRA compared with PPD include the following:

  • Only a single patient visit required
  • Ex vivo tests
  • No booster effect
  • Independent of BCG vaccination

Disadvantages of IGRA include the following:

  • High cost
  • More laboratory resources required
  • Complicated process of lymphocyte separation
  • Lack of prospective studies

Screening test accuracy

Neither tuberculin skin testing nor IGRA testing is sufficiently sensitive to rule out TB infection. [67] Approximately 20% of patients with active TB, particularly those with advanced disease, have normal PPD test results. In addition, caution is recommended on the interpretation of these tests in infants and in patients with immunosuppressive conditions. [65]

A systematic review of QuantiFERON-TB Gold (QFT-G)/Gold in-Tube (QFT-GIT) and T-SPOT.TB by Chang and Leung concluded that QFT-G had the highest positive likelihood ratio (48.1) for latent TB infection and that T-SPOT.TB had the best negative likelihood ratio (0.10). A negative T-SPOT.TB result in middle-aged and older patients makes active TB very unlikely. [68]

Results from a study by Leung et al indicated that tuberculin skin testing is not predictive of the subsequent development of active TB. [69] The authors followed 308 men with increased risk of TB because of silicosis. A positive T-SPOT.TB finding was associated with a relative risk of 4.5 for subsequent TB in the group overall and a relative risk of 8.5 among the men who did not receive preventive treatment for latent TB. The CFP-10 spot count was more predictive than the ESAT-6 spot count.

In a study by Diel et al, all subjects who developed active TB within 4 years after exposure to a smear-positive index case had positive results using QFT-GIT. These researchers concluded that QFT-GIT was more reliable than tuberculin skin testing for identifying patients, especially children, who will soon progress to active TB. [70]

In a study of kidney-transplant recipients, positive ELISpot assays predicted the subsequent development of TB in patients who did not have a significant tuberculin skin test reaction or risk factors for TB infection. Active TB developed after kidney transplantation in 4 of 71 patients (6%) with positive ELISpot assays, but in none of the 201 patients with negative or indeterminate ELISpot results [71]