TB: Bronchoscopy or sputum collection?

Excluding tuberculosis

The situation

Excluding the possibility of tuberculosis is a common clinical situation. Older patients, chronic lung disease and increased use of immunosuppression increase the risk of tuberculosis reactivation and have created an increasingly common scenario where active contagious tuberculosis needs to be excluded quickly and with minimum inconvenience.

Is a single bronchoscopic specimen good enough?

Why is it so hard?

Tuberculosis is also often the great mimicker. It can present as many other chronic lung conditions ranging from pneumonia to emphysema. It always requires a high level of suspicion to sufficiently exclude it. Making its exclusion purely on the basis of history, clinical examination or radiographs very difficult. Skin testing has no role in the exclusion of current active disease (they are best at predicting future reactivation)

“Ruling out” tuberculosis can be both time consuming (sputum collection) and inconvenient (isolation) and stressful to the patient’s family.

Why do we have to be so sure?

It is also necessary to exclude tuberculosis with a higher degree of certainty than other diseases due to the public health contagion issues that come with tuberculosis. This may be even greater for those patients referred from group housing situations such as nursing homes where a larger number of patients are at risk.

So what is the optimal way to exclude tuberculosis?

Some of my colleagues suggested that a single bronchoscopy specimen with a negative AFB is sufficient to exclude contagion. I can see why this is attractive. It is a single step procedure performed in a highly controlled setting with a high sample collection rate (as opposed to sputum collection and induction). A smear is quickly run and if negative the patient can be safely removed from isolation and even sent home. But is all this backed by good science?

To answer the issue I did a review of the literature. The following are my findings.

Anderson et al conducted a study where consecutive patients were subjected to hypertonic saline sputum induction followed by fiberoptic bronchoscopy. The results were as follows:

Method Sensitivity NPV
Bronchoscopy 73% 91%
Sputum collection 87% 96%

Where NPV is negative predictive value

Sputum collection was better than bronchoscopy for excluding tuberculosis in this series. 1

Brown et al in 2007 compared sputum induction with gastric washing to bronchoscopically obtained specimens. They were able to show that consecutive sputum collection was superior to single bronchoscopy collected specimen.2

The above studies were referenced in the CDC Guidelines for the preventing the transmission of Mycobacterium tuberculosis which specifically gives the following recommendations under the bronchoscopy sub heading:3

For AFB smear and culture results, three sputum samples have an increased yield compared with a single specimen (110,357), and induced specimens have better yield than specimens obtained without induction. Sputum induction is well-tolerated (90,109,132, 133,357,361,362), even in children (134,356), and sputum specimens (either spontaneous or induced) should be obtained in all cases before a bronchoscopy (109,356,363,364).

In circumstances where a person who is suspected of having TB disease is not on a standard antituberculosis treatment regimen and the sputum smear results (possibly including induced specimens) are negative and a reasonably high suspicion for TB disease remains, additional consideration to initiate treatment for TB disease should be given. If the underlying cause of a radiographic abnormality remains unknown, additional evaluation with bronchoscopy might be indicated; however, in cases where TB disease remains a diagnostic possibility, initiation of a standard antituberculosis treatment regimen for a period before bronchoscopy might reduce the risk for transmission. Bronchoscopy might be valuable in establishing the diagnosis; in addition, a positive culture result can be both of clinical and public health importance to obtain drug-susceptibility results. Bronchoscopy in patients with suspected or confirmed TB disease should not be undertaken until after consideration of the risks for transmission of M. tuberculosis (30,63,81,162,360). If bronchoscopy is performed, because it is a cough-inducing procedure, additional sputum samples for AFB smear and culture should be collected after the procedure to increase the diagnostic yield.


Bronchoscopic sputum collection should not be done instead of induced sputum collection. If samples are collected by bronchoscopy further samples should be collected by sputum induction to maximize yield.


  1. Anderson, C, N Inhaber, and D Menzies. 1995. Comparison of sputum induction with fiber-optic bronchoscopy in the diagnosis of tuberculosis. American journal of respiratory and critical care medicine 152, no. 5 Pt 1 (November): 1570-4. http://www.ncbi.nlm.nih.gov/pubmed/7582296
  2. Brown, Michael, Hansa Varia, Paul Bassett, Robert N Davidson, Robert Wall, and Geoffrey Pasvol. 2007. Prospective study of sputum induction, gastric washing, and bronchoalveolar lavage for the diagnosis of pulmonary tuberculosis in patients who are unable to expectorate. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 44, no. 11 (June): 1415-20. doi:10.1086/516782. http://www.ncbi.nlm.nih.gov/pubmed/17479935
  3. Jensen, P. 2005. Guidelines for Preventing the Transmission of Mycobacterium tuberculosis in Health-Care Settings, 2005. MMWR. http://www.cdc.gov/mmwr/preview/mmwrhtml/rr5417a1.htm

Related studies

[mendeley type=”collections” id=”10378561″ sortorder=”authors”]


One thought on “TB: Bronchoscopy or sputum collection?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.