Limits of the current diagnostic approach for COVID-19
The diagnostic gold standard for COVID-19 is the naso-pharingeal (NP) swab reverse-transcription real-time polymerase chain reaction (rRT-PCR) detection of SARS-Cov-2. However, due to the unavailability of a shared reference standard for COVID-19 diagnosis there are no reliable data on NP swab sensitivity. Clerici et al. assessed nasopharyngeal swab sensitivity in patients with known SARS-Cov-2 infection, based on the presence of symptoms and of ≥1 positive rRT-PCR serial testing, and found a sensitivity of 77% (95% CI, 73 to 81%).6Wang et al. evaluated SARS-Cov-2 detectability in different biological specimens in COVID-19 patients and found a NP swab sensitivity of 63%.7 Pooled data found that the probability of a false negative result was as high as 21% even at the optimal testing window (3 days after symptom onset).8
Given the limits of NP swab testing, some experts propose to diagnose suspected cases using the widely available, time-saving and non-invasive imaging approach of chest computed tomography (CT), that could serve as an efficient and effective way to flag, diagnose, and possibly triage COVID-19 patients.9,10 However, as confirmed by a recent metanalysis of 60 studies (5744 patients), CT has a low specificity compared to NP swabs rRT-PCR, 46% (95% CI, 29-63%).11 Ongoing studies are evaluating the role of radiomics analysis to identify a diagnostic signature for COVID-19 infection, based on standard-of-care chest CT imaging, with promising preliminary results showing a sensitivity of 69.52%, and a specificity of 91.63%.12
In this scenario identifying the false negative cases remains of critical importance to properly manage patients avoiding improper allocation of COVID-19 cases and allowing a timely treatment. Since the early pandemic, BAL has been proposed as a potentially useful diagnostic tool to increase COVID-19 diagnosis sensitivity. Nevertheless, considering the high potential to aerosol exposure generate during BAL, international bronchology societies have universally cautioned for a limited and proper use of this tool in clinical practice during the pandemic peaks. The role of BAL in the diagnostic algorithm of COVID-19 has been debated and explored in several studies.