We have read with great interest the original article entitled "Comparative analysis of chest radiography and lung ultrasound to predict in-hospital prognosis in patients admitted for SARS-CoV-2 pneumonia (COVID-19)" by Rubio-Gracia et al.1 We would like to congratulate the authors, as it highlights the value of clinical ultrasound in the management of pathologies with pulmonary involvement.
It highlights the use of a high sensitivity test2,3 to evaluate potentially infectious patients, and raises the comparison with a technique of questionable diagnostic accuracy. Although we agree that the comparison with the reference test, computed tomography (CT), would have allowed us to draw more important conclusions, its implementation is subject to logistical constraints, which were even greater in March 2020, with a shortage of human, material and space resources,4 not to mention the radiation and high costs involved in its implementation as a screening tool.
Promoting the use of lung ultrasound in these settings may improve outcomes, not only in detecting clinical deterioration, but also to indicate isolation precautions in early stage patients.2,3 It is a technique that can easily be applied in the context of major structural changes, such as those that occurred during the SARS-CoV-2 pandemic.
Chest radiography, which is widely used, has significant limitations in the detection of involvement,2 despite being the first tool used.
Regarding their results,1 the poor correlation with conventional radiography may be influenced by the score used, as the proposed scale was described for patients admitted to critical care units, which does not fit the sample used in their study.
On the other hand, the mean age of the patients requiring hospital admission is striking in the results. This may have led to a lack of significance in one of the variables that determine an unfavourable outcome.
Based on the patients who were included in our working group,5 we would like to contribute to their study by suggesting other interesting risk factors to be considered.
By looking at more accessible variables, such as age and other easily accessible variables, like C-reactive protein, we were able to generate a predictive model that could help predict hospital admission and mortality. Under the name "rule of 7",5 we were able to obtain a sensitivity and specificity of 56.8% and 87.6% and an area under the curve (AUC) of 0.813, with a value of p < 0.001, for mortality. For this, a C-reactive protein value greater than 70 mg/L, an age above 70 years and a LUS > 7 were associated.
However, despite ample evidence supporting the value of the technique in SARS-CoV-2 involvement, standardised staging to grade the severity of lung involvement and to establish a common language has yet to be determined.
In any case, we are grateful that the use of techniques that are safe for both the patient and the practitioner, and no less cost-effective, is highlighted.
FundingThis work has received no public subsidies or financial support. No funding sources were used to assist in the preparation of this study. This research did not receive any specific grants from funding bodies in the public, private or non-profit sectors.
Authors' contributionAll authors read and approved the final manuscript. All authors have contributed to this paper.
Conflict of interestThe authors declare that they have no conflict of interest in relation to this article.
