Article Text


Research letter
Tuberculosis through the rose tinted spectacles of the EBUS endoscopist: be aware of the bias
  1. Kurt Tournoy
  1. Correspondence to Dr Kurt Tournoy, Ghent University Hospital, Department of Respiratory Medicine, Building 7 K12 IE, De Pintelaan 185, Ghent 9000, Belgium; kurt.tournoy{at}

Statistics from

I read with interest the article on the utility of endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) in tuberculous intrathoracic lymphadenopathy by Navani et al.1

EBUS-TBNA has been validated for the assessment of mediastinal nodes in lung cancer2 and to obtain a diagnosis in (presumed) centrally located lung cancer3 or sarcoidosis.4 In addition to a recent report,5 the study by Navani et al adds to the evidence for the use of EBUS-TBNA in cases of presumed tuberculous lymphadenitis. A sensitivity of 94% is reported, which might be too optimistic.

First, patients were selected in a peculiar way. The authors reviewed the files of all EBUS endoscopies and retrospectively selected those cases in which tuberculosis was finally found. Unfortunately, there is no information on how the patients were selected beforehand. The reported figure gives an indication of the sensitivity of EBUS in this particular setting; however, it does not give an answer to the more relevant question about the sensitivity of EBUS-TBNA for all cases in whom tuberculous lymphadenitis is suspected. There were potentially many patients with tuberculous intrathoracic lymphadenitis who were not sent for EBUS.

Second, the use of assessment tools (ie, EBUS) only in patients having the condition leads to an overestimation of sensitivity. Since there is no remedy for the overestimation in this series, the results should be interpreted with caution.

Finally, three of the five pathology grades are grouped as compatible with tuberculosis. Two of these, epithelioid granulomas without caseation and necrosis are primarily compatible with sarcoidosis and cancer rather than tuberculosis,4 despite suggestive symptomatology or an (undefined) response to medication. A more conservative analysis combining strict pathological and microbiological criteria would be informative.

Therefore, it might be appropriate to say that for tuberculous lymphadenitis, the sensitivity of EBUS is at the most 94%. Although I recognise the importance of EBUS, my reflections should serve as a reminder to doctors to exercise caution when their diagnosis of tuberculosis is based on the idea that the sensitivity of EBUS is 94% and that a negative EBUS excludes the disease.


View Abstract


  • Competing interests None.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • PostScript
    Neal Navani Onn Min Kon Sam M Janes