Jump to Main Content

COVID-19 Resources

COVID-19 and VTE/Anticoagulation: Frequently Asked Questions

(Version 14.0; last updated February 2, 2022)

Input from Drs. Lisa Baumann Kreuziger, MD; Agnes Y. Y. Lee, MD, MSc; David Garcia, MD; Maria DeSancho, MD; and Jean M. Connors, MD.

Note: Please review ASH's disclaimer regarding the use of the following information.

ASH has convened a multidisciplinary, international panel to develop guidelines on the use of anticoagulation in patients with COVID-19. These guidelines are “living” in that the updates are made as data become available. In collaboration with experts on the guideline panel, the team conducts urgent systematic reviews of available evidence to make evidence-based recommendations. When available, ASH recommendations will be included in this FAQ.

We all note that several authors of this FAQ have participated as investigators in clinical trials relevant to our recommendations, and thus may have intellectual conflict-of-interests.

Is COVID-19 associated with an increased risk for venous thromboembolism (VTE)?

The incidence of VTE in COVID-19 patients varies depending on the patient population. In a meta-analysis of 66 observational studies through August 2020, the overall prevalence of VTE in hospitalized patients was 9.5 percent without screening ultrasound (US), and 40 percent with screening US; and higher in intensive-care-unit (ICU) patients at 18.7 percent without, and 45.6 percent with US. As with other medical patients, those with more severe disease, especially in the setting of additional risk factors (e.g., age, being male, obesity, cancer, history of VTE, comorbid diseases, ICU care), have a higher risk of VTE than those with mild or asymptomatic disease. It is important to note that these data were gathered prior to Omicron and Delta outbreaks, and the relative risk of VTE with these strains is unknown.

When objective imaging is not immediately feasible in a COVID-19 patient to confirm or refute a diagnosis of pulmonary embolism (PE) or deep vein thrombosis (DVT), medical providers must rely on clinical assessment based on history, physical findings and other tests. Likelihood of PE is moderate to high in those with signs or symptoms of DVT, unexplained hypotension or tachycardia, unexplained worsening respiratory status, or traditional risk factors for thrombosis (e.g., history of thrombosis, cancer, hormonal therapy). The value of D-dimer testing is the ability to effectively rule out PE/DVT when the level is normal, given low false negative rates. It is important to realize that D-dimer levels are higher in patients with COVID-19, especially those with severe or critical disease, independent of the presence or absence of PE/VTE, and thus generally cannot be used to imply a diagnosis of VTE/PE. The presence of high D-dimer levels have been shown to be associated with more severe COVID-19. Elevated levels have been used in some trials of anticoagulants to risk stratify patients. Whether elevated D-dimer level alone predicts response is unclear, with trials showing variable results (see below).

If a patient with COVID-19 requires therapeutic anticoagulation for VTE or AFIB stroke prevention, are there any special considerations to consider, such as drug interactions?

 Multiple medications are being used for COVID-19 treatment. Dexamethasone is an inducer of CYP3A4 and the extent of the drug interaction with direct oral anticoagulants is unknown. Rivaroxaban and apixaban have significant drug interaction with ritonavir, a component of the antiviral PAXLOVID (nirmatrelvir 300 mg with ritonavir 100 mg). Co-administration will increase the concentration of apixaban or rivaroxaban and may increase the risk for bleeding. As PAXLOVID treatment lasts only five days, the clinical significance of the interaction with DOACs remains unclear. Alternatives to PAXLOVID in high-risk non-hospitalized patients with COVID-19 include sotrovimab, remdesivir, or molnupiravir. The potential risks and benefits of using PAXLOVID in an individual patient on rivaroxaban/apixaban must be weighed in the context of whether sotrovimab is available as an alternative COVID-19 treatment and on the patient’s susceptibility to severe COVID-19 disease. Remdesivir is believed to have no clinical effect on metabolic pathways. Monoclonal antibodies should not affect DOAC or warfarin metabolism. The University of Liverpool has collated a list of drug interactions. low-molecular-weight heparin (LMWH) or UFH in hospitalized critically ill patients is preferred because of the shorter half-life and fewer drug-drug interactions compared with direct oral anticoagulants. Regular warfarin users who are unable to get INR monitoring during isolation may be candidates for direct oral anticoagulant therapy, but patients with mechanical heart valves, ventricular assist devices, valvular atrial fibrillation, antiphospholipid antibody syndrome, or lactation should continue treatment with warfarin therapy. LMWH or UFH remain the anticoagulants of choice in pregnancy.

What dose of anticoagulation should be empirically used in hospitalized COVID-19 patients in the absence of confirmed or suspected VTE?

All hospitalized adults with COVID-19 should at a minimum receive pharmacologic thromboprophylaxis, unless the risk of bleeding even on prophylactic dosing outweighs the risk of thrombosis. LMWH is preferred over unfractionated heparin (UFH). In the setting of heparin-induced thrombocytopenia, fondaparinux is recommended. In patients for whom anticoagulants are contraindicated or unavailable, mechanical thromboprophylaxis (e.g., pneumatic compression devices) can be used. Combined pharmacologic and mechanical prophylaxis is not generally recommended.

Optimal anticoagulant dosing to reduce COVID-19 complications has been actively investigated. Multiple randomized controlled trials (RCTs) have been initiated and some have reported results. For the purposes of these trials, critically ill patients are usually defined as those requiring ICU level care such as hemodynamic support, ventilatory support, and/or renal replacement therapy; and moderately ill as those admitted to hospital but not requiring ICU-level care; however, definitions can vary between trials. Outcome endpoints have differed between trials, and need to be kept in mind when reviewing results.

The largest RCT for hospitalized patients, a multiplatform adaptive-design trial (MPT), incorporated three multicenter studies/networks (REMAP-CAP, ATTACC, and ACTIV-4A; N Engl J Med 2021;385:777-789; N Engl J Med 2021;385:790-802) to address the question of whether therapeutic anticoagulation is indicated in critically ill or moderately ill patients. The primary outcome was a composite of 21-day “organ support–free” days, defined as the number of hospital days not requiring high-flow nasal oxygen, invasive or noninvasive mechanical ventilation, vasopressor therapy, extracorporeal membrane oxygenation (ECMO) support together with in-hospital mortality. Thrombosis, bleeding, and overall mortality were secondary outcomes. Patients who required therapeutic anticoagulation for other indications were excluded. The INSPIRATION trial compared intermediate dose versus standard prophylactic dose LMWH in critically ill patients. The HEP-COVID trial compared therapeutic with prophylactic/intermediate dose LMWH anticoagulation in hospitalized patients with D-dimer level >4× the upper limit of normal (ULN) or sepsis-induced coagulopathy score of greater than 4. The RAPID trial compared therapeutic with prophylactic doses of heparin in moderately ill hospitalized patients with elevated D-dimer level. Additional RCTs have been reported at meetings or on pre-print servers or are ongoing.

For patients requiring ICU or ICU level of care:

The MPT initially assessed therapeutic dose versus institutional prophylaxis regimens (including both standard prophylactic and intermediate doses of heparin) in 1,098 critically ill patients. The trial was stopped early per pre-specified criteria for futility, with no benefit for therapeutic anticoagulation in reducing need for organ support or death at 21 days. There was also no difference in survival to discharge. It should be noted that approximately 50 percent of patients in the standard of care arm received intermediate dose heparin, and 22.4 percent of patients in the therapeutic dose arm did not receive this dose. Major bleeding was seen in 3.8 percent of those assigned to therapeutic dose heparin and in 2.3 percent of those assigned to standard of care. For thrombotic events, the rate was 6.4 percent in the therapeutic dose arm and 10.6 percent in the standard prophylaxis arm; however, thrombosis alone was not a predefined secondary outcome.

The INSPIRATION trial examined intermediate dose LMWH compared with standard prophylactic dose in ICU patients with a primary composite endpoint of venous or arterial thrombosis, ECMO, or mortality within 30 days. The results of this RCT found no benefit with intermediate dose compared to standard prophylactic dose, with major bleeding 2.5 percent in the intermediate dose arm and 1.4 percent in the standard prophylactic dose arm.

Based on these data, we advise that critically ill patients receive standard prophylactic doses of anticoagulants, since increased doses of heparin do not confer a benefit for preventing progression of COVID-19 or death. The ASH guideline panel currently suggests using prophylactic-intensity over intermediate-intensity or therapeutic-intensity anticoagulation in patients with COVID-19–related critical illness who do not have suspected or confirmed VTE (conditional recommendation based on very low certainty in the evidence about effects).

For hospitalized patients not requiring ICU level of care:

In hospitalized moderately ill COVID-19 patients (n=2,219), the MPT was stopped early for superiority, finding benefit for the use of therapeutic dose anticoagulation compared to institutional standard thromboprophylaxis in prevention of disease progression, based on the same primary composite outcome used for the critically ill cohort. 28.2 percent assigned to the standard care arm received higher than prophylactic dose heparin and 20.3 percent assigned to therapeutic dose heparin received a lower dose, complicating interpretation. There was a 4 percent benefit of freedom from organ support or death at 21 days for those receiving therapeutic dose anticoagulation. There was no difference in survival to discharge. Major bleeding occurred in 1.9 percent of those assigned to therapeutic anticoagulation compared to 0.9 percent in those assigned to standard care dosing. Major thrombotic events occurred in 1.1 percent of patients in the therapeutic dose arm and 2.1 percent in the standard care arm.

The RAPID trial randomized 465 non-ICU hospitalized patients with elevated D-dimer to therapeutic dose or prophylactic dose heparin; intermediate dosing was not allowed. The primary outcome was a composite of death, invasive mechanical ventilation, non-invasive mechanical ventilation, or ICU admission, and was not statistically significantly different between the two arms (OR, 0.69; 95% CI, 0.43-1.10; p=0.12), although the odds of death at 28 days was decreased in the therapeutic dose arm (OR, 0.22; 95% CI, 0.07-0.65; p=0.006). There was a numeric decrease in the number of VTEs, with two events in the therapeutic dose arm and seven events in the standard prophylactic dose arm. There was no difference in bleeding.

The HEP-COVID trial randomized 253 hospitalized patients deemed to be at increased risk based on an elevated D-dimer > 4× ULN or an elevated sepsis-induced coagulopathy score to therapeutic versus prophylactic dose LMWH and found a significant 13.2 percent absolute risk reduction (28.7% vs. 41.9%) in the composite of arterial and venous thrombosis and all-cause mortality in the therapeutic dose arm, driven by decreased thrombotic events, as there was no difference in deaths between the two groups. This benefit was found in the non-ICU/moderately ill patients only (67.2% of the trial population), and not those that required ICU level of care (32.8%). Overall, there were six major bleeds in the therapeutic dose arm (2 in non-ICU stratum and 4 in ICU stratum) and two major bleeds in the standard of care (non-ICU stratum).

As in the post-discharge anticoagulation discussion above, the ACTION trial comparing therapeutic dose rivaroxaban for moderately ill inpatients followed by 30 days post discharge did not decrease events or mortality at 30 days compared to inpatient prophylactic dose of LMWH only.

Although the benefit of therapeutic heparin was shown in the three RCTs, the trial inclusion criteria and definitions of positive outcomes differed. Differences in control group anticoagulant intensity occurred across trials, and significant proportions of patients receiving anticoagulant doses higher or lower than the assigned treatment complicates interpretation. There was no difference in survival to discharge in the MPT trial enrolling the largest number of patients, and the absolute risk difference for progression to organ support or death at 21 days was only 4 percent, with a 1 percent major bleeding rate. This benefit may or may not be considered compelling for an individual patient. However, identifying which patients would benefit is challenging, as the studies within the MPT had different inclusion criteria regarding time from hospital admission to enrollment. There appears to be a signal for decreased VTE in these trials for moderately ill patients with full-dose anticoagulation; however, VTE was not a stand-alone prespecified outcome in most trials. The HEP-COVID and RAPID trials required an elevated D-dimer, suggesting that D-dimer might stratify patients at increased risk; although in the MPT, D-dimer in the moderately ill did not predict a differential treatment effect of therapeutic dose anticoagulation.

Based on review of the above data, the ASH draft guideline panel suggests using therapeutic-intensity over prophylactic-intensity anticoagulation for patients with COVID-19–related acute illness who do not have suspected or confirmed VTE or another indication for anticoagulation (conditional recommendation based on very low certainty in the evidence about effects). The NIH COVID-19 Guideline panel also recommends using therapeutic-dose heparin for patients who have a D-dimer above the upper limit of normal (ULN), require low-flow oxygen, and have no increased bleeding risk.

It is important to note that clinical trials available to date and utilized in the guidelines processes were performed prior to Omicron and most instances prior to Delta strain outbreaks. Given apparent differences in Omicron severity and target organs, it is possible that current results may not apply, and more research is urgently needed. It is also important to point out that these recommendations apply only to patients hospitalized due to COVID-19 disease, not to those discovered incidentally to have asymptomatic or mild SARS-CoV-2 infection on routine screening at the time of hospital admission for another indication, as is currently very common for many patients found to be infected with the Omicron strain.

How should we manage COVID-19 patients who experience recurrent clotting of access devices (e.g., central venous catheters, arterial lines) or extracorporeal circuits (e.g., CRRT, ECMO) despite prophylactic anticoagulation?

Although of unproven benefit, it may be reasonable to increase the intensity of anticoagulation (i.e., from standard-intensity prophylaxis to intermediate-intensity prophylaxis or from intermediate-intensity prophylaxis to therapeutic-intensity) or switch anticoagulants in these settings. Any decision to increase the intensity of anticoagulation should take into account the individual patient’s bleeding risk.

Should ambulatory or post-hospital discharge patients with COVID-19 receive anticoagulation for prophylaxis?

We advise not starting anticoagulation for acutely ill COVID-19–positive outpatients. Patients diagnosed with mild-to-moderate COVID-19 not requiring hospitalization do not benefit from starting anticoagulation, either as thromboprophylaxis or to prevent progression of COVID-19, based on a very low event rate in stable outpatients, from the ACTIV-4b RCT comparing placebo, aspirin, or two different doses of apixaban in ambulatory patients older than 40 years.

The risk of VTE following hospital discharge appears low and similar to risks following hospital discharge for other medical conditions, based on observational studies. The ASH Clinical Practice Guidelines do not recommend VTE prophylaxis in this post-discharge population. Two RCTs evaluating post discharge VTE prophylaxis have been subsequently reported. The ACTION trial published that use of therapeutic dose rivaroxaban 20 mg daily during hospitalization followed by continued post-discharge use for 30 days resulted in no benefit compared to in-hospital prophylactic dose LMWH. The efficacy of inpatient versus post-discharge anticoagulation cannot be determined because outcomes were only reported at 30 days after discharge. The MICHELLE trial showed a significant reduction in post-discharge thrombotic events (including screen-detected DVT and PE at day 35) and death after 35 days of treatment with rivaroxaban 10 mg per day compared to no anticoagulation in a population of patients selected for increased VTE risk based on the IMPROVE-DD score.

We advise no routine anticoagulation for post discharge patients. However, an assessment can be performed to determine risk for VTE, using scores such as IMPROVEDD, to determine whether an individual patient may merit post-discharge prophylaxis, also considering bleeding risk and feasibility. For example, patients with COVID-19 who are discharged early to free up inpatient beds (“home hospital” approach) might still have significantly reduced mobility and merit continued thromboprophylaxis. All patients with COVID-19 should be educated on the signs and symptoms of VTE at hospital discharge and advised to seek urgent medical attention should these develop. Enrollment in clinical trials to address this question is encouraged.

Is the risk of VTE increased following COVID-19 vaccination?

As described in a separate FAQ, patients receiving the adenoviral AstraZeneca or Johnson &Johnson vaccines are at small but real risk for vaccine-induced immune thrombosis and thrombocytopenia (VITT), a form of thrombosis and thrombocytopenia syndrome (TTS). However, despite case reports documenting VTE in the absence of VITT following COVID-19 vaccination, larger observational and epidemiologic studies as well as analysis of the original large phase III mRNA vaccine registration trials have not confirmed an overall increased risk for VTE following COVID-19 vaccination with mRNA or adenoviral vaccines. We would not recommend special monitoring or prophylaxis for VTE post-vaccination..

For additional information, see:


ASH thanks Menaka Pai, MD, MSc; Menno Huisman, MD; Stephan Moll, MD; Mary Cushman, MD, and Walter Ageno, MD for their contributions to earlier versions of this FAQ, and to FAQs focused on COVID coagulopathy and DIC and pulmonary embolism that have been retired, with content incorporated into the current version of this FAQ.

View All COVID-19 FAQs