BAL in COVID-19 in the critically ill patients.
Several studies have evaluated the utility of BAL in the critically ill patients, two were prospective.18,19,33-36 A summary of BAL findings in critically ill patients is reported in Table 1 and 2. The highest positivity for SARS-Cov-2 detection in BAL performed in critically ill patients has been reported by Wang W et al, 93% (95%CI 074-1.00; N total BAL = 15) and Yang Y et al, 68% (95% CI 056-0.79; N total BAL = 44).7,19,37 The latter study reported a 100% SARS-Cov-2 positivity in the more severe patients in whom BAL was collected within the first 2 weeks. After 15 days the positivity od nasopharyngeal and oropharyngeal swabs decreased while BAL maintained a high positive rate of 63%.37 Gao et al designed a retrospective study to evaluate the diagnostic accuracy of nasopharyngeal swab (NP) compared to BAL for the detection of SARS-Cov-2.38 They reviewed 123 intubated patients who underwent both tests (time interval median 1day, IQr 1-2.75 days) showing that 9 cases with negative NS had positive BAL, 7% of the total. The remaining cases were: 70 positive for both, 39 negatives for both, and 5 cases with positive NS and negative BAL. Bacterial pneumonia was identified in 34% of total cases, with significantly more bacterial coinfections in the non-COVID-19 (24/44, 54%) than in the COVID patients (18/79, 23%).38 Similar results were achieved by Mahmood et al in 55 critically ill patients, in the subgroup of 37 negative NS they found one positive BAL for SARS-Cov-2 (3%) and in the overall cohort found 16% of positive cultures other than COVID-19 (Staphylococcus, Pseudomonas, Fungi, Mycobacterium avium and Pneumocysist jirovecii ).39 In the ICU setting BAL allows the detection of coinfections in a significant proportion of COVID-19 (Table 2). In several studies conducted in the ICU setting BAL was mainly performed for a microbiological purpose, with a significant impact in subsequent medical decisions. Baron et al performed BAL in 24 patients for microbiological purposes and only in 2 (7%) BAL was performed to confirm COVID-19 after negative NS.40 The Authors describe the use of BAL mainly for a suspicion of ventilator associated pneumonia (N=11, 39%), invasive aspergillosis (N=4, 14%) and to rule out superinfection before starting a steroid course. In this study, BAL had an impact on medical decisions in 20 cases (71%), with introduction (n = 6), continuation (n = 3), switch (n = 2), or withdrawal (n = 4) of antimicrobial therapy in 14 cases (50%) and/or decision to start (n = 6; 21%), or not (n = 6, 21%), corticosteroid therapy.40 Pickens et al conducted a retrospective single centre study in COVID-19 mechanically ventilated patients, documenting by early BAL (48h within intubation) 21% (28/133) of bacterial superinfection pneumonia. Streptococcus species and methicillin-susceptible S. aureus (MSSA) combined accounted for 79% (22/28) of cases.33 Polymicrobial infections were common, three patients, previously treated with antibiotics had pathogens resistant to standard CAP antibiotics—oneStenotrophomonas maltophilia and two methicillin-resistantS. aureus (MRSA) and Pneumocystis was found in one patient with HIV on antiretroviral treatment. For each day of mechanical ventilation they measured the Narrow Antibiotic Treatment (NAT) score and found a clinically and statistically significant difference between positive and negative BAL results (NAT score median difference -1, 95% CI -1 to 0; p=0.001). These findings suggest that negative BAL analysis was used to narrow or discontinue antibiotic treatment and that in the absence of BAL ventilator associated pneumonia may be underrecognized yet overtreated with unnecessary broad antibiotics.33