Finding an Evidence-Based and Clinically Important Role for BAL in the Setting of Suspected SARS-Cov-2 Infection.

To the Editor:

We read with interest the study by Hamed et al in this issue of CHEST that compares nasopharyngeal swabs (NPS) and BAL for the detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The authors demonstrate a significant viral gradient from the upper to the lower respiratory tract and a significantly higher sensitivity with the BAL than with the NPS (96% vs 67%). They conclude that a BAL should be obtained in the absence of an existing positive result for SARS-CoV-2.

Given the lack of studies comparing the sensitivity of bronchoscopy and less invasive methods, , this is certainly a most welcome research. However, the retrospective design, nonconsecutive enrollment, relatively small sample size, and extreme specificity of the study population deserve mention. The study cohort, in particular, is composed exclusively by critically ill patients (86.5% intubated or on extracorporeal membrane oxygenation) with SARS-CoV-2 infection (100% disease prevalence). It is very likely that pretest clinical probability of disease in this population was extremely high. Doubts remain on the reproducibility of the results in a cohort of patients in which SARS-CoV-2 infection was only one of the possible diagnoses. Furthermore, the clinical utility of invasive testing in patients with a very high pretest probability of SARS-CoV-2 infection is lower, because a negative test might not lead to a significant change in the patient’s treatment.

In the only other comparative study, a BAL was performed within 48 hours of at least one negative NPS in 79 patients with hypoxemic respiratory failure whose condition did not require intubation. A 97.5% agreement between the two tests was observed, and only two patients with a negative NPS were diagnosed with coronavirus disease 2019 on the BAL. Although the authors do not specify the final diagnosis in patients with negative NPS and BAL for SARS-CoV-2 infection, it is likely that the prevalence of coronavirus disease 2019 in this population was very low, making it completely different from that enrolled by Hamed et al.

In conclusion, the two aforementioned studies underline the critical importance of the population being examined. It is key that BAL and less invasive methods be compared prospectively in a cohort of consecutive patients with suspected SARS-Cov-2 infection who have been enrolled based on criteria decided beforehand, preferably across a wide spectrum of disease severity. This would allow us to decide reliably when it is clinically useful to perform an invasive procedure that, in this specific setting, implies organizational complexity and risks to the health-care staff.