Sensitivity and feasibility of lung ultrasound in bronchiolitis -- reply to the correspondence letter by Catalano

We thank Dr Catalano et al. for their interest on our article. The comment raises two important issues, one regarding the test sensitivity and the second, more general, on test feasibility. Assuming the final diagnosis as a gold standard, lung ultrasound (LUS) showed an excellent sensitivity (47/52 vs 38/52 of chest x-ray, 90 vs 73 %), but we have to consider the observational characteristics of the study design, performed in a real-world setting with real patients, real problems, and real doctors, with clinically driven indication to chest xrays. This is typical of an observational effectiveness study; it is not an efficacy study, evaluating the technique under ideal conditions, and the observed results can be more directly relevant to clinical practice but also vulnerable to selection bias [1]. The authors raise concerns about the equipment setting and rationale for chest x-ray prescription. We reported this information in the Methods section: chest x-ray was always requested by the attending physician on the basis of the clinical picture; the equipment used for the lung ultrasound was a Toshiba Nemio machine, equipped with a high-resolution linear probe with frequencies ranging from 6 to 12 MHz. Regarding the second issue on the practical appeal of the method, we reported our gratifying experience in the demanding theater of a neonatal intensive care unit, consistent with the international consensus that the method is simple, safe, and sensitive—when challenged under controlled conditions versus appropriate gold standards [5]. Dr Catalano argues that lung ultrasound relying on artifacts, such as B-lines, is not an evidence-based best-practice imaging clue. We agree with that: LUS is not always able to provide a detailed anatomical definition, especially when we consider B-lines. However, we believe that the usefulness of LUS relies in being an additional imaging tool for the evaluation of the lung, which is quick, bedside, low-cost, non-ionizing, and very sensitive to detect the loss of air content. Of course, LUS should be integrated with other imaging modalities, when a detailed anatomical definition is needed. It is also true that every new diagnostic tool always raises many uncertainties when it is proposed. Given the high versatility and potential clinical implications of adding ultrasound to the regular imaging equipment, we believe it is of importance to deepen the knowledge of this technique to better understand its strengths and limitations. Cardiologists who experienced lung ultrasound as an adjunct to much more technically demanding echocardiography say that “from a technical viewpoint, in the echocardiographic cursus studiorum where 2D echo represents the elementary school, Doppler echo the secondary school, and stress echo University, B-lines correspond to kindergarten [3].” We also concur with Daniel Lichtenstein, who recently stated (Montecarlo Lung Ultrasound School 2013) that it takes 20 s to make a lung scan with V. A. Caiulo (*) School of Ultrasound of Italian Federation of Pediatricians, Piazza Angeli 3, 72100 Brindisi, Italy e-mail: antoniocaiulo@inwind.it