Response.

To the Editor:

We thank Nadeem et al for taking interest in our study that reports the benefit of high-dose prophylactic anticoagulation (HPA) on thrombotic complications in critically ill patients with COVID-19.

Nadeem et al hypothesized that patients treated with standard thromboprophylaxis were more severe and had more chest scans, thus more thrombotic complications. This possibility calls for several comments: (1) To evaluate the effect of cumulative individual exposure specifically to HPA on thrombotic complications, we used an original method based on a dedicated time-varying exposure model. Cumulative individual exposure was used as a surrogate for “time within the therapeutic range” and allowed a more appropriate evaluation of the anticoagulation regimen, exposed to frequent changes of doses in ICU. (2) According to the Groupe d’Intérêt en Hémostase Périopératoire guidance document, the anticoagulant dose increased with the severity of COVID-19 pneumonia. (3) The benefit of HPA on thrombotic complications remained significant after adjustment for severity markers that included SOFA score, Pao 2/Fio 2 ratio, renal replacement therapy, and extracorporeal membrane oxygenation status. In our study, renal replacement therapy was performed in the ICU with trained staff, and renal replacement therapy filter clotting was recorded as an event only when it appeared unusual to experienced physicians.

We agree with Nadeem et al that we may have underestimated the complications of bleeding, and this point is addressed in the discussion.

The current challenge is to identify accurately the patients who are at particularly high thrombotic risk who could benefit from HPA and then determine when the period of high thrombotic risk ends to prevent bleeding events by switching to standard thromboprophylaxis. Biomarkers may play a crucial role to monitor this biphasic progression; during the initial inflammatory phase, when thrombotic complications are reported, prothrombotic and inflammatory markers (which include D-dimers, fibrin monomers, fibrinogen, CRP, and IL-6) reach very high levels. , After a few days in the ICU, they gradually decrease, along with the inflammatory syndrome and with a potential increase in bleeding risk. To identify this shift between thrombotic risk and hemorrhagic risk, we need to monitor a set of biomarkers that reflect the evolution of thromboinflammation. Patient profile, disease progression, and biomarkers could be combined to modulate thromboprophylaxis strategy to capture the complex interaction between thrombotic and bleeding risks.