Air-splinting maneuver during flexible fiber-optic bronchoscopy.

Correspondence TB is appropriate, particularly in high-prevalence countries. The study by Lee et al 2 performed in a high-prevalence Asian region and involving a very high number of untreated contacts demonstrates that the risk of progression is at its highest directly after Mycobacterium tuberculosis (MTB) infection and that more than one-half of those subjects who progressed to TB did so within the fi rst 2 years. Thus, the statistical power of a progression study will be greatest when the period of high risk of progression is studied, that is, the fi rst couple of years after exposure to infectious patients. Furthermore, the problem of participant dropout will be minimized if the focus is on the fi rst few years following recent infection. From randomized studies on the preventive treatment of latent TB infection, such as the large-scale International Unit Against Tuberculosis Eastern European study published in 1982, 3 we learned that following infection, there is a natural decline in the risk of developing TB over time. Therefore, we agree with Drs Khurana and Khurana that a follow-up of people for 2 years at a given date makes little sense if infection with MTB occurred decades earlier (eg, in early childhood), unless there is impaired immunocompetence. Without any knowledge of the specifi c conditions of a previous MTB infection among the people tested, a positive result, irrespective of whether a tuberculin skin test (TST) or an interferong release assay (IGRA) is used, cannot, per se, lead to the conclusion that there is a high risk of the person developing TB in the future. Thus, it follows that decisions on chemoprevention should not be based on IGRA or TST results alone. As neither the IGRA nor the TST distinguishes old from recent infection, the history of exposure needs to be analyzed to decide whether old or recent infection is likely. This also pertains to the screening of health-care workers. The high prevalence of positive TST or IGRA results and the low rate of progression for health-care workers indicate that old infections with a low risk of progression prevail. 4 Of particular note is the fact that in the small study (n 5 48) by Park et al, 5 a considerable variation in IGRA results is observed, especially with concentrations around the cutoff. Again in agreement with Drs Khurana and Khurana, the meaning of these phenomena needs to be elucidated by generating data on risk of progression depending on IGRA variation. 4