Ambulatory Patients with COVID-19
For ambulatory (non-hospitalized) adult patients with COVID-19, the overall risk of thromboembolism does not appear to be markedly elevated.[9, 10] The National Institutes of Health (NIH)-sponsored ACTIV-4b clinical trial comparing placebo, aspirin, and two doses of apixaban for patients with COVID-19 infection who did not require hospitalization was recently stopped due to a low number of thromboembolic events across all treatment groups. Overall rates of death, symptomatic arterial or venous thromboembolism, myocardial infarction, stroke, or cardiopulmonary hospitalization were extremely low in all groups (0.7–1.4%) without any meaningful difference between groups.
For patients who already take chronic antithrombotic therapy (e.g., stroke prevention in atrial fibrillation, secondary prevention of coronary artery disease), observational studies have shown mixed results. Some have found an association between use of antithrombotic agents and improved outcomes while others have failed to demonstrate such a benefit.[12, 13] Nonetheless, even in the presence of COVID-19 infection, these patients usually retain their indication for antithrombotic therapy. Therefore, unless their clinical condition changes and the risk of bleeding significantly increases, continued use of the chronic antithrombotic agent is generally advised.
We recommend against any specific antithrombotic preventative therapy for ambulatory (non-hospitalized) adult patients with mild COVID-19 infection who have no other indication for antithrombotic therapy.
We recommend patients on antithrombotic therapy prior to diagnosis of COVID-19 continue their antithrombotic therapy unless a significantly elevated risk of bleeding has developed.
Non-Critically Ill Hospitalized Patients with COVID-19
For adult patients who are not critically ill but still hospitalized for COVID-19 and do not otherwise have an indication for anticoagulation, four completed clinical trials have explored the potential role of different intensities of anticoagulation (online appendix). In the multiplatform clinical trials (ATACC, ACTIV-4 A, REMAP-CAP) of more than 2200 patients, use of therapeutic-intensity heparin (primarily enoxaparin) was superior to ‘usual care’ pharmacologic thromboprophylaxis (71.7% standard dose, 26.5% intermediate dose) with heparin (primarily low-molecular-weight heparin [LMWH]) for increasing organ support-free days which is the number of days without cardiovascular (e.g. use of vasopressor or inotropic medications) or respiratory (e.g. use of high-flow nasal cannula, noninvasive or invasive mechanical ventilation) support (80.2% vs. 76.4%, adjusted difference 4.0%, 95% CI 0.5–7.2%). Presence of elevated baseline D-dimer did not significantly alter the primary outcome. Overall survival until hospital discharge was not different among the two treatment arms (adjusted risk difference 1.3%, 95% CI − 1.1 to 3.2%), however there was a reduction in the secondary outcome of major thromboembolic events or death (2.6%, 95% CI 0.2–4.4%) favoring therapeutic heparin. There was a numerically higher risk of major bleeding with treatment-dose heparin (1.9% vs. 0.9%) that did not reach statistical significance. Of note, these trials screened more than 13,000 patients. Key strengths of this trial are the large size, global site inclusion, and use of blinded adjudication of event outcomes. But interpreting clinicians should be aware that > 12,000 of the screened patients were not included and that approximately one-quarter of patients received an intermediate-dose of heparin. This may have blunted any differences in efficacy and safety outcomes between the two treatment groups. Furthermore, interpretation of these absolute risk differences should be done carefully given the adaptive, Bayesian nature of the study design rather than traditional 1:1 randomization as is often seen in most traditional randomized clinical trials. Patients at high risk of bleeding were excluded from the study, which may partially explain the low rate of major bleeding in the trial population. The definition of ‘major’ thrombotic events did not include deep vein thrombosis; however, the addition of deep vein thrombosis did not alter the results of the secondary outcome analysis. There is also some question as to the value patients place on the outcome of the absolute number of organ support-free days versus any need for organ support, objective thromboembolic events, or overall survival. Finally, while large and well conducted, there are concerns about the relatively low rate of symptomatic thromboembolic events and lack of mortality benefit.
In the RAPID COVID COAG study, 465 moderately ill adult patients admitted with COVID-19 and elevated D-dimer levels were randomized to receive therapeutic intensity or standard dose prophylactic heparin. Greater than 90% of both treatment arms received LMWH (vs. unfractionated heparin). Patients were included if their D-dimer level was > 2-times the hospital upper limit of normal (ULN) or if it was > 1-time the upper limit of normal along with documented hypoxia. The primary endpoint of death, invasive or non-invasive mechanical ventilation, or ICU admission was not statistically different between the two groups (16.2% versus 21.9% for therapeutic and standard intensity prophylaxis, respectively, OR 0.69, 95% CI 0.43–1.10). Interestingly, the secondary outcome of all-cause mortality was reduced with therapeutic intensity prophylaxis (1.8% vs. 7.6%, OR 0.22, 95% CI 0.07–0.65) while the VTE risk was low and similar in the two groups (0.9% vs. 2.5% in the therapeutic and prophylactic groups, respectively, OR 0.34, 95% CI 0.07–1.71). The rate of major bleeding was similar in both groups (0.9% vs. 1.7%, OR 0.52, 95% CI 0.09–2.85). VTE risk was low and similar in the two groups − 0.9% versus 2.5% in the therapeutic and prophylactic groups, respectively (OR 0.34, 95% CI 0.07–1.71). This trial screened nearly 4000 patients to enroll and randomize 465 patients. Strengths of this trial include the use of blinded adjudication of outcomes and a set of study endpoints that are patient centric. However, RAPID COVID COAG was likely underpowered to detect a difference in key outcome measures.
In the HEP-COVID trial, 257 patients admitted to the hospital with COVID-19 and D-dimer levels > 4-times the ULN were randomized to receive standard dose or therapeutic intensity thromboprophylaxis with heparin (mostly LMWH). Patients were stratified at the time of randomization into ICU (32.8%) and non-ICU (67.2%) cohorts. Use of therapeutic intensity thromboprophylaxis reduced the primary efficacy outcome (arterial or venous thromboembolism or death) significantly (28.7% versus 41.9%, RR 0.68, 95% CI 0.49–0.96), which was driven by the non-ICU cohort (16.7% versus 36.1%, RR 0.46, 95% CI 0.27–0.81). Major bleeding was non-significantly elevated in patients receiving therapeutic intensity thromboprophylaxis (4.7% versus 1.6%, RR 2.88, 95% CI 0.59–14.02) overall but rates were similar among the non-ICU stratum (2.4% versus 2.3%). Of note, this trial screened more than 11,000 patients to enroll 257, largely excluding patients who did not meet criteria or lacked sufficiently elevated D-dimer levels, which raises concerns about external validity. Furthermore, this study relied on local adjudication of clinical events, which can lead to biased assessments when treatment allocation is known. The analysis also included asymptomatic VTE events identified on routine lower extremity compression ultrasound study at hospital day 10 ± 4 (or hospital discharge), the clinical importance of which are unknown. However, symptomatic DVT was significantly reduced (5.4% vs. 15.3%, RR 0.35, 95% CI 0.15–0.81) while symptomatic VTE was non-significantly reduced (1.6% vs. 2.4%, RR 0.38, 95% CI 0.12–1.19).
In the ACTION trial, 615 patients randomized to treatment-dose rivaroxaban experienced similar risk of death, duration of hospitalization, and duration of supplemental oxygen as compared to prophylactic dose heparin. The composite thrombotic outcome (VTE, myocardial infarction, stroke, and major adverse limb events) was similar in the two treatment groups (7% vs. 10%, RR 0.75, 95% CI 0.45–1.26). The outcome of ISTH-defined major bleeding was higher in the rivaroxaban group than with prophylactic dose heparin (8% vs. 2%, RR 3.64, 95% CI 1.61–8.27).
A few key points can be abstracted from these four published trials. First, use of therapeutic intensity non-heparin thromboprophylaxis (e.g., rivaroxaban) did not demonstrate benefit over standard dose thromboprophylaxis with heparin and increased the risk of major bleeding. Second, each of the randomized trials used a different D-dimer threshold for inclusion and/or analysis subgroups. Third, each of these trials had strict inclusion and exclusion criteria that selected for low bleeding risk patients, like most anticoagulant trials for a new indication, suggesting that therapeutic intensity anticoagulation, if used, should be applied only after weighing risks and benefits.
We recommend that all patients hospitalized with COVID-19 receive at least standard dose thromboprophylaxis.
In patients admitted to the hospital for indications other than COVID-19 but incidentally found to have COVID-19 infection, we recommend standard dose thromboprophylaxis with LMWH or unfractionated heparin (UFH) unless specific contraindications exist.
We suggest that clinicians consider the use of therapeutic intensity LMWH or UFH thromboprophylaxis for non-critically ill patients at increased risk of disease progression or thromboembolism and who are not high risk for anticoagulant-related bleeding (Table 2).
We recommend that “intermediate” intensity thromboprophylaxis and/or antiplatelet agents only be used in the setting of a clinical trial for hospitalized patients with COVID-19.
Given that there are no head-to-head comparisons of various heparin agents, use of either LMWH or UFH can be considered for thromboprophylaxis. However, most heparin used in the clinical trials was LMWH which also affords less exposure and personal protective equipment utilization.
Suggested UFH and LMWH regimens for hospitalized non-pregnant adults, including adjustments for renal impairment and obesity, are reflected in Table 3.
Whenever therapeutic intensity of UFH, LMWH, or other anticoagulants are given for thromboprophylaxis and not for treatment of presumed or confirmed thromboembolism, this should be clearly documented in the patient chart.
Critically Ill Hospitalized Patients with COVID-19
Three studies have examined the use of non-standard doses of heparin for thromboprophylaxis in critically ill adult patients with COVID-19 (online appendix). In the multi-platform trials, patients admitted to the ICU and requiring organ support (e.g., high-flow oxygen, invasive ventilation, vasopressor or inotropic support) did not experience benefit when receiving therapeutic intensity thromboprophylaxis (primarily LMWH) as compared to ‘usual care’ thromboprophylaxis (primarily LMWH). Of note, use of “intermediate” intensity thromboprophylaxis was quite common (51.7%) in the ‘usual care’ thromboprophylaxis group. Two trials have evaluated intermediate versus standard dose thromboprophylaxis in this population and failed to find a statistically significant benefit. In the INSPIRATION trial, patients admitted to the ICU did not experience benefit when receiving intermediate-dose enoxaparin (1 mg/kg daily) as compared to standard dose thromboprophylactic enoxaparin (40 mg daily). Finally, a small, multi-center randomized trial of 176 patients with COVID-19 admitted to the ICU found a non-significant reduction in 30-day all-cause mortality associated with intermediate-dose enoxaparin as compared to standard dose thromboprophylactic enoxaparin (15% vs. 21%, OR 0.66, 95% CI 0.30–1.45) with a low rate of major bleeding (2% in each arm)..
Of note, approximately one-third of patients in the previously described HEP-COVID study required nonrebreather oxygen mask or more intensive respiratory support. The primary efficacy endpoint (a composite of arterial thromboembolism, VTE, or death) was not significantly reduced in the stratum of patients admitted to the ICU (51.1% vs. 55.3%, RR 0.92, 95% CI 0.62–1.39).
We recommend that adult patients who are critically ill at the time of hospitalization receive standard dose thromboprophylaxis instead of intermediate- or therapeutic intensity thromboprophylaxis.
Suggested dosing regimens for hospitalized non-pregnant adults, including adjustments for renal impairment and obesity, are reflected in Table 3.
Two trials have explored the use of antiplatelet therapy in patients hospitalized with COVID-19. The RECOVERY trial enrolled adult patients in the United Kingdom, Indonesia, and Nepal who were hospitalized with COVID-19. Patients were randomized to receive aspirin 150 mg daily or usual care. Among the 14,892 patients who were randomized, the rate of 28-day all-cause mortality was similar in the two groups (17% vs. 17%, RR 0.96, 95% CI 0.89–1.04). There was also no significant difference in the median time to discharge alive (8 vs. 9 days). In the ACTIVE-4a study, 562 non-critically ill patients hospitalized for COVID-19 in Brazil, Italy, Spain, and the United States were randomized to receive therapeutic intensity heparin plus a P2Y12 inhibitor (ticagrelor preferred) or therapeutic intensity heparin alone. The primary outcome of organ support-free days was similar in both treatment groups (21 vs. 21, adjusted OR 0.83, 95% CI 0.55–1.25). There was also no difference in the rate of survival to hospital discharge or major bleeding.
that patients hospitalized with COVID-19 do not receive antiplatelet therapy (e.g., aspirin, P2Y12 inhibitor) for the specific purpose of preventing thromboembolism or COVID-19 disease progression.
Hospital Transitions of Care
Many patients who are hospitalized for COVID-19 infection will require transfers of care between the floor and ICU and vice versa. No studies have specifically compared different anticoagulation strategies when patients initially admitted to one unit (e.g., floor) require a chance in level of care (e.g., transfer to an ICU). The study protocols of the various randomized trials generally recommended patients remain on their initial intensity of anticoagulation even when transferring between different levels of care (e.g., floor to ICU) for up to 14 days or until recovery since the patients were enrolled and randomized shortly after hospital admission.[14, 18] Any decisions to use higher intensity anticoagulation should be clearly documented to provide guidance on approaches to post ICU management. Furthermore, there is no prospective clinical evidence to support the use of serial D-dimer testing to guide the intensity of antithrombotic therapy.
We suggest that adult patients admitted to the hospital for COVID-19 remain on the intensity of VTE thromboprophylaxis that was initiated at hospital admission as long as their bleeding risk is not significantly elevated.
Thus, a moderately-ill patient with COVID-19 admitted to the ward and started on therapeutic intensity thromboprophylaxis should continue therapeutic intensity thromboprophylaxis when transferring to the ICU. The dose should be reduced to standard dose thromboprophylaxis when clinically necessary based on bleeding risk.
Similarly, patients initially admitted to the ICU for organ support and started on standard dose thromboprophylaxis should continue standard dose thromboprophylaxis when they transfer to the floor.
Patients should receive therapeutic intensity anticoagulation if a thromboembolic event is confirmed or highly suspected, and bleeding risk is not prohibitively high.
Observational studies have reported conflicting results regarding post-hospital VTE risk. In most reports, the observed risk of VTE was similar to patients without COVID-19.[9, 23,24,25] However, in one report, the risk of post-hospital VTE was elevated compared to patients without COVID-19. Most recently, the MICHELLE trial randomized 320 patients hospitalized for COVID-19 who were receiving standard dose thromboprophylaxis during their admission and were considered at increased risk for post-discharge events, with an IMPROVE VTE score of ≥ 4 or an IMPROVE VTE score of 2–3 plus a D-dimer > 500 ng/ml at discharge. Patients at increased risk of bleeding, such as those with bleeding in prior 3 months, on dual antiplatelet therapy or with chronic kidney disease, were excluded. Only 11 patients with creatinine clearance 30–50 ml/min were included in the study. Enrolled subjects were randomized to rivaroxaban 10 mg by mouth daily for 35 days or no post-discharge thromboprophylaxis. Post-hospital thromboprophylaxis with rivaroxaban reduced the primary composite outcome of symptomatic venous or arterial thromboembolism, VTE-related death, bilateral VTE, myocardial infarction, non-hemorrhagic stroke, major adverse limb event or cardiovascular death compared to no intervention (3.14% VS. 9.43%, RR 0.33, 95% CI 0.13–0.90). There were no major bleeds in either group. However, it should be noted that the MICHELLE trial screened 997 patients in order to enroll 320, suggesting that post-hospital extended thromboprophylaxis is not appropriate for all patients hospitalized with COVID-19. Note that the IMPROVE VTE score used in the MICHELLE trial gave 1 point for immobilization (confined to bed or chair with or without bathroom privileges) that lasted ≥ 1 day. This contrasts with the original IMPROVE VTE score which defined immobilization as ≥ 7 days immediately prior to and including hospitalization..
We recommend that clinicians not routinely use post-hospital thromboprophylaxis after discharge following hospitalization for COVID 19 for all patients, including those who may have received therapeutic intensity anticoagulation for thromboprophylaxis.
We suggest post-hospital thromboprophylaxis with rivaroxaban 10 mg daily for 35 days following a hospitalization for COVID-19 may be considered in select patients at increased risk of thromboembolism (e.g., IMPROVE VTE score ≥ 4 or score 2–3 with elevated D-dimer at hospital discharge) and not at increased risk of bleeding regardless of the intensity of their inpatient thromboprophylaxis.
We recommend clear documentation and communication of indication and intended duration of post-hospital thromboprophylaxis to providers and next care settings to avoid unnecessarily prolonged exposure to anticoagulation.