| 899 with ARDS (16%) without any specific aPL profile. Thus, it is difficult to draw conclusions based on so few aPL-positive patients. Because of the lack of a standardized assay and aPL heterogeneity, their detection is dependent on enzyme-linked immunosorbent assays (2), which were provided by different manufacturers in the 2 studies. In particular, coating or buffer composition can influence the interaction between cardiolipin and cofactor proteins (3), and can influence whether antibodies are detected or not. Moreover, knowing the difficulties of interpreting aPL positivity, we pay close attention to their routine detection. All positive samples are controlled in duplicate, and, to address the specificity, the absorbance of uncoated wells treated in the same conditions is subtracted in order to avoid nonspecific binding (4). In contrast to Frapard and colleagues who focused on patients in the ICU, we also included patients with severe COVID-19 who were not admitted to the ICU and observed that 29% were positive for IgG aCL (Figure 1A). There was no difference in IgG aCL levels between aCL-positive patients admitted versus those not admitted to the ICU (Figure 1B), suggesting that IgG aCL positivity in severe COVID-19 is not attributable to the pulmonary complications alone.If the link between viral infection and aPL detection is well described in the literature (5), the relationship between aPL positivity and lung injury is not obvious. To support their hypothesis, Frapard et al referenced a study by Wiedermann et al (6), in which an association between aPL positivity and lung injury was not shown except in 1 patient with catastrophic antiphospholipid syndrome (CAPS). In the literature, the association between aPL positivity and ARDS is rarely described and only reported in the context of APS or CAPS (7). Consistent with our results, Zuo et al ( 8) recently reported that half of patients hospitalized for