Our systematic review and meta-analysis of 14 studies involving 7,681 Covid-19 patients provides the a comprehensive examination of the risks of bleeding, thrombosis and death for patients with and without anticoagulation treatment, and patients underwent different types of anticoagulation treatment (prophylactic or therapeutic). The current results might indicates that prophylactic or therapeutic anticoagulation were superior to no anticoagulation in reducing the mortality.
Persistent evidence shown that pharmacological thromboprophylaxis can significantly reduce the risk of venous thrombus embolism for general patients 35. However, the role of prophylactic anticoagulation in Covid-19 patients is unclear. Some studies revealed the potential benefits of anticoagulant treatment in severe Covid-19 patients with higher venous thrombus embolism risk 36–38, such as reducing the mortality risk, while some studies indicated that routine chemical prophylaxis was inadequate in preventing venous thrombus embolism in severe Covid-19 patients 39. In our study, anticoagulation treatment, no matter prophylactic or therapeutic, showed a mild effectiveness of reducing mortality among Covid-19 patients (Prophylactic anticoagulation vs No anticoagulation: OR = 0.80, 95%CI: 0.69–0.93; Anticoagulation vs No anticoagulation: OR = 0.91, 95%CI: 0.80–1.05). The results were consistent with current guidelines advocating treatment and the use of prophylactic and/or therapeutic anticoagulants in patients with Covid-19 appeared to be advocated 40. According to the current studies, the mechanism of anticoagulant therapy could reduce the thrombosis in patients with Covid-19 might be as follows 41,42. After SARS-CoV-2 causes infection, the virus would attack angiotensin converting enzyme 2 (ACE2) and decrease the ACE2 content, the role of ACE2 being to convert angiotensin II to angiotensin 1 and 7. Therefore, a decrease in ACE2 leaded to an increase in angiotensin II and a decrease in angiotensin 1 and 7, and these changes in angiotensin levels leaded to an increase in superoxide levels. Recruitment of neutrophils leaded to an increase in superoxide production and an increase in superoxide production leads to endothelial cell dysfunction through the nitric oxide pathway. Because there were vesicles containing von Willebrand Factor (vWF) in endothelial cells, when a 500% increase in its activity, these vesicles exulted, and through a number of complex interactions, an increase in vWF could lead to an increase in local thrombus where the inflammation occurs.
Preventive and therapeutic anticoagulant therapy might have opposite clinical effectiveness of reducing bleeding and thrombosis for Covid-19 patients. Venous thromboembolic disease has been reported as one of the major complications occurring in patients with Covid-19 43. Even in patients with Covid-19, there was currently no clear guidance on anticoagulant dose. Jimenez et al believed that patients could benefit from intermediate anticoagulation dosages26. Bleeding and thrombus were the two extremes of coagulation dysfunction 44. Our results suggested that the risk of bleeding was greater with therapeutic anticoagulation than with prophylactic anticoagulation, while the risk of thrombus was greater with prophylactic anticoagulation than with therapeutic anticoagulation, which may be related to the treatment dose 45,46. The results suggested the need to pay attention to drug dosage in anticoagulant therapy, and the need for higher level of evidence-based medicine. Therefore, the therapeutic dose should be carefully considered in clinical practice.
The literatures included in this study were not stratified according to severity of the disease, as a result, we could not evaluate the clinical benefits for mild and severe patients separately. But, when the severity of the disease was corrected as a confounding factor in a multivariate regression model, higher doses were found to be beneficial for the prevention of death, indicating that the effectiveness value was consistent across different disease degree stratification29. Another study showed a much higher frequency of pulmonary embolism in ICU patients with Covid-19 (21%) than during the same time interval in 2019 (6%), and it was also higher than the incidence of pulmonary embolism in patients with influenza admitted to the same ICU in 2019 (8%) 47. It seemed clear therefore that the incidence of pulmonary embolism in patients admitted to ICU with Covid-19 was much higher than in other critically ill non-Covid-19 patients, including those with acute respiratry distress syndrme and other respiratory infections, despite the fact that these patients are already at an increased risk of pulmonary embolism 48. In the studies mentioned above 39, as was the case in the present study, patients developed pulmonary embolism even though most of them were receiving anticoagulant thromboprophylaxis. These findings raise the need of that the dose of anticoagulant therapy should be carefully considered in the process of clinical diagnosis and treatment, particularly in higher risk Covid-19 patients 6,49,50.
Limitations
Our study has several limitations. First, in the outcome event, subgroup analysis usually consisted of only one literature, for which we grouped each subtype into a single broad category, including bleeding events, thrombosis events, and death events. Second, this study was conducted when the disease outbreak is ongoing. Many regions affected by Covid-19 haven not yet published clinical datasets, which may skew the results of this analysis. All these datasets are retrospective, which prevents us from exploring risk factors. Additionally, due to the limitation of sample size, our analysis could not be divided into subgroups such as region and dose. Finally, the meta-analysis was performed by statistical result data, therefore there was no way to analyze case data according to more detailed clinical needs.