There is more than one way to catch a bug.

veolar lavage fluid or mononuclear cells [4, 5] . Single serologic antibody detection tests for TB lack sensitivity, but might be promising when used as multiple-antigen tests in combination with other procedures [6] . There are many other diagnostic tools under investigation, such as volatile organic compounds, gold nanoparticle probes, biomarkers in QuantiFERON supernatants for example. However, none of the above-mentioned strategies reaches sufficient diagnostic yield to replace sputum smear microscopy and TB culture. In this issue of Respiration , Lee et al. [7] propose another approach for the rapid diagnosis of sputum smearnegative pulmonary TB. In their retrospective analysis of 84 patients with confirmed active pulmonary TB and with a negative sputum smear, they compared the diagnostic value of clinical findings, chest high-resolution (HR) CT, sputum TB-PCR and the IFNrelease assay (IGRA) for the diagnosis of active pulmonary TB. As has been shown previously, they found none of these parameters alone to be of sufficient diagnostic value. However, they describe a high positive and negative predictive value for the combined utilization of IGRA and chest HRCT. A congruent result of HR-CT and IGRA, i.e. HR-CT findings suggestive of pulmonary TB and a positive IGRA, predicted the correct diagnosis in 23 of 24 patients (positive predictive value 96%). The negative predictive value of congruent negative HR-CT findings and IGRA was 92%: in 23 of 25 patients pulmonary TB was correctTuberculosis (TB) kills. In 2007 there were an estimated 9.27 million new TB cases worldwide and about 1.78 million people died from the disease in the same year. Sputum smear microscopy remains the primary tool for TB diagnosis in most countries, but only an estimated 44% of TB cases are smear positive [1] . Though culture of TB bacteria is more sensitive, it may take several weeks to diagnosis due to low growth rates. Research is ongoing to develop the best diagnostic strategy for smear-negative cases (still 5.2 million people after all), with ‘best’ in this context implying highly sensitive, specific, fast, reliable and, last but not least, cost effective. Recently published papers on the diagnosis of smearnegative pulmonary TB focus on the improvement in clinical judgment, specimen collection or specimen processing, and innovative laboratory technologies. A score based on demographic data, symptoms, past medical history and chest X-ray findings for rapid identification of patients with pulmonary TB reached a sensitivity of 93%, but had a very low specificity (42%) questioning its clinical significance [2] . New technologies such as LED fluorescence microscopy and bleach microscopy might improve sensitivity of smear microscopy [3] . In developed countries, flexible bronchoscopy in patients unable to produce sputum or with smear-negative sputum might accelerate the diagnostic procedure, especially when TBspecific nucleic acid amplification techniques or enzymelinked immunospot assays are performed on bronchoalPublished online: March 23, 2010