While Waiting to Buy a Ferrari, Do Not Leave Your Current Car in the Garage!

to buy EBUS-TBNA, given the period of limited economic resources, the better bet is to use conventional TBNA in a rational way, in order to try and keep the diagnostic success rate high, while containing the costs. In this view, studies like the one by Medford et al. [6] published in this issue of Respiration are certainly welcome. They analyze performance characteristics and cost utility of TBNA in two cohorts of patients with different pre-test clinical probability of malignancy, [6] . Over the 2-year study period, the authors tried, deliberately, to select two different study populations: in the 1st year, in fact, they enrolled a clinically unselected cohort of patients with enlarged mediastinal nodes (38 patients, prevalence of malignancy: 71%), whereas in the 2nd year they selected only patients with a high pre-test clinical probability of malignancy (41 patients, prevalence of malignancy: 95%). Patients with a negative TBNA were systematically submitted to mediastinoscopy, unless they were unfit for surgery and had a reliable diagnosis of malignancy obtained with other sampling methods (e.g. bronchial biopsy or transthoracic needle aspiration). Overall sensitivity and accuracy of TBNA were high (78 and 84%, respectively), especially in the group with a high prevalence of malignancy (86 and 87%, respectively). Mediastinoscopy could be avoided in 25% of patients (up to 37% of patients in the high-prevalence cohort). According to the authors’ calculations, their diagnostic apThe possibility to sample the mediastinum does probably represent the most important progress associated with the practice of diagnostic bronchoscopy in the last 25 years. Transbronchial needle aspiration (TBNA) is a well-recognized, cheap and safe procedure, which has spared a huge number of patients the costs and risks of unnecessary diagnostic or therapeutic surgical procedures [1] . More recently, the swift development of health care technologies has provided the interventional pulmonologists with the opportunity to overcome the only great flaw of TBNA, i.e. its ‘blindness’. Real-time endobronchial ultrasound-guided TBNA (EBUS-TBNA), in fact, allows the examiner to localize with exquisite detail the mediastinal structures, and to sample the lymph nodes with great accuracy, while avoiding the great vessels [1] . There is a robust and growing body of literature demonstrating that EBUS-TBNA offers substantial diagnostic advantage over conventional TBNA, especially for small lymph nodes ( ! 1.5 cm) and specific lymph node stations (e.g. 2 R-L and 4L) [2, 3] . These characteristics make EBUS-TBNA significantly better than conventional TBNA as a method for mediastinal staging of lung cancer in patients with a low prevalence of lymph node metastases, which basically include most candidates for surgery [3–5] . Unfortunately, however, EBUS-TBNA is quite expensive in capital and running costs, and its availability is not yet widespread. For those who still can just dream Published online: February 17, 2010