A simple, reproducible and accurate lung ultrasound technique for COVID-19: when less is more.

We appreciate the interest of Mento et al. in our study and their interesting work on lung ultrasound (LUS) [1]. We also thank these authors for the opportunity to further discuss LUS in coronavirus disease 2019 (COVID-19).

The authors state that we have “limited the inspected areas to six”. However, the LUS technique used in our study was performed by scanning 10 areas covering the whole anterior, lateral and posterior chest bilaterally, including the two lung bases, and not six [2]. This is well detailed in the Methods paragraph and previously published papers [3, 4]. Thus, we fully agree that all the visible lung parenchyma should be scanned in patients suspected of COVID-19, which is exactly what was done in our study.

When we designed our protocol, the challenge was to propose a highly feasible technique and reproducible pattern recognition to several operators from different countries and continents with variable skills during the real-life experience of a pandemic surge, and not only to few experts from the same hospital. Our data demonstrate that this LUS approach is feasible and reproducible, while more complicated techniques and inconsistency could have increased unnecessarily the risk of variability. Moreover, our data demonstrate that the intermediate probability LUS patterns efficiently differentiated a lower probability of real-time polymerase chain reaction (RT-PCR) positivity in comparison to the high probability. The potential undervaluation supposed by Mento et al. wouldn’t have had any practical deleterious consequences in the decision flow chart and wouldn’t have changed the very high sensitivity of LUS [2].

Our study was not designed to perform a quantitative analysis “in the management of patients affected by COVID-19”, but rather to investigate the potential of LUS in the first diagnosis of patients suspected of COVID-19. Indeed, while triaging patients undoubtedly represents a highly impactful application of LUS targeted to speed up the process of patient’s allocation in a crowded hospital, a complex quantitative assessment by chest imaging is still of uncertain practical usefulness as well as of doubtful generalized applicability in emergency.

We appreciate the technique suggested by Mento et al. claimed as a “proposal for international standardization” (reference 4 in their letter), but some strong limitations should be highlighted. (A) The “internipple” line does not divide the lung into two halves, superior and inferior. In most patients in the supine position, there is no more than a small part of the visible lung or even just abdomen below this anatomical line (Fig. 1). (B) The separation in a high number of areas that do not change the extension of the examined chest, complicates the exam and the risk is to discourage the potential operators. (C) The indication to use a wireless device with a second operator who remotely reads the images by a tablet further complicates the exam and is not needed to reduce the possibility of cross-infections.

Fig. 1

CT scans of two patients with COVID-19 pneumonia: the arrows indicate the internipple line, demonstrating that this anatomic landmark does not separate the lung into two halves, superior and inferior. This is particularly true if we consider that the CT scan is usually done with the arms of the patient over the head, that dislocates the nipples in a superior position

In conclusion, we agree that LUS in COVID-19 pneumonia needs a rigorous approach over the entire chest, as we have implemented in our multicenter study; however, it is crucial to avoid unnecessary complications that do not add any real clinical value.