Version July 2025
INTRODUCTION —
Deep vein thrombosis (DVT) and pulmonary embolism (PE) are forms of venous thromboembolism (VTE). VTE has significant morbidity and mortality for patients in the community and in hospital. Anticoagulation is the mainstay of therapy for VTE. The purpose of anticoagulation is the prevention of recurrent thrombosis, embolization, and death, the risk of which is greatest in the first three to six months following the diagnosis.
Following initial anticoagulation for the first 5 to 10 days, patients with VTE require therapy for a more prolonged period. This topic review will discuss the selection of an anticoagulant, the transition from initial therapy, and the duration and monitoring of treatment phase (long-term) anticoagulation. The indications for anticoagulation, an overview of DVT treatment, and details regarding initial and indefinite anticoagulation for patients with VTE are discussed separately. (See "Overview of the treatment of lower extremity deep vein thrombosis (DVT)" and "Venous thromboembolism: Initiation of anticoagulation" and "Selecting adult patients with lower extremity deep venous thrombosis and pulmonary embolism for indefinite anticoagulation".)
The approach to anticoagulation outlined in this topic is, in general, consistent with strategies outlined by several international societies including The American College of Chest Physicians, The American College of Physicians, The European Society of Cardiology, The European Respiratory Society, The American Society of Hematology, and others [1-5].
TERMINOLOGY —
For the purposes of this topic, the following terms apply (table 1) [6]:
●Provoked DVT – A provoked DVT is one that is precipitated by a known event (eg, surgery, hospital admission, estrogen). It is now also referred to as DVT associated with an identifiable risk factor(s) (table 2).
Risk factors can be transient or persistent. Transient risk factors are further categorized as "major" or "minor" according to the magnitude of VTE risk conferred. The role of risk factors in selecting patients for indefinite anticoagulation is discussed separately. (See "Selecting adult patients with lower extremity deep venous thrombosis and pulmonary embolism for indefinite anticoagulation", section on 'Characterize risk factors'.)
●Unprovoked DVT – An unprovoked DVT is one that occurs in the absence of identifiable risk factors.
●Proximal DVT – A proximal DVT is located in the popliteal, femoral, or iliac veins; or the inferior vena cava (figure 1).
●Distal DVT – An isolated distal DVT (also known as isolated calf vein DVT) has no proximal component, is located below the knee, and is confined to the calf veins (peroneal, posterior, anterior tibial, and muscular veins). The popliteal vein is not involved.
●Phases of anticoagulation – Anticoagulant therapy for acute DVT is often divided into three phases.
•The initiation phase (or initial therapy) uses parenteral or higher-dose oral agents to control acutely propagating thrombus and prevent extension or embolization.
•The maintenance phase (also called "treatment phase" or "long-term") of anticoagulation therapy is administered for a finite period beyond initiation, usually three to six months and occasionally up to 12 months. The goal of this phase is stabilizing the thrombus while recovery via intrinsic thrombolysis takes place.
•The extended phase (indefinite) of anticoagulation therapy is administered beyond the finite period, sometimes indefinitely. The goal of this phase is secondary prevention.
●Direct factor Xa and thrombin inhibitors – Direct factor Xa and thrombin inhibitors have been referred to as newer/novel oral anticoagulants, non-vitamin K antagonist oral anticoagulants, direct oral anticoagulants, and target-specific oral anticoagulants [7,8]. Throughout this topic, we refer to these agents by their pharmacologic class, direct factor Xa and thrombin inhibitors, consistent with society recommendations [8]. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects".)
INDICATIONS —
Details of the indications for anticoagulation for DVT and PE are discussed separately. (See "Overview of the treatment of lower extremity deep vein thrombosis (DVT)", section on 'Indications for anticoagulation (most patients)' and "Acute pulmonary embolism in adults: Treatment overview and prognosis", section on 'Definitive therapy'.)
SELECTION OF AGENT —
Treatment phase (long-term) anticoagulant therapy is administered beyond the initial few days of anticoagulation for a finite period of typically three to six months and up to 12 months in some cases (ie, scheduled stop date). Options for treatment phase anticoagulation include oral anticoagulants (ie, factor Xa inhibitors, direct thrombin inhibitors, and warfarin) and parenteral subcutaneous anticoagulants (low molecular weight [LMW] heparin and fondaparinux). While the oral factor Xa and thrombin inhibitors are typically preferred, choosing among these options frequently depends upon clinician experience and availability, the risks of bleeding, patient comorbidities and preferences, cost, and convenience (table 3).
Patient values and preferences are critical in selecting a treatment phase agent for anticoagulation in acute VTE. While most patients prefer oral agents, oral factor Xa inhibitors and oral thrombin inhibitors are generally much more costly than warfarin, but there is a lower burden from laboratory monitoring, dose adjustment, and drug-drug and drug-food interactions. Some patients may express a preference for daily injections rather than weekly international normalized ratio (INR) monitoring and other burdens of warfarin [9]. As another example, patients who place a high value on cost may choose warfarin while others who place a high value on a lower risk of bleeding may prefer the direct oral anticoagulants (DOACs). (See "Selecting adult patients with lower extremity deep venous thrombosis and pulmonary embolism for indefinite anticoagulation", section on 'Patient values and preferences'.)
General population — As a general principle, oral agents are typically preferable to parenteral agents. For most nonpregnant patients, we suggest the DOACs, rivaroxaban, apixaban, edoxaban, or dabigatran, rather than warfarin (table 4) and suggest warfarin rather than LMW heparin. Factors that influence agent selection including feasibility of oral administration are discussed below (table 3).
●Factor Xa inhibitors (eg, rivaroxaban, apixaban, edoxaban) and oral direct thrombin inhibitor (ie, dabigatran) – These oral agents are our preferred anticoagulant for most hemodynamically stable, nonpregnant, nonlactating patients who do not have severe kidney insufficiency (table 4). While rivaroxaban and apixaban can be administered as monotherapy (a higher dose is used during the initiation phase), edoxaban and dabigatran are administered following a five-day course of heparin for initiation phase therapy. Guidelines do not favor a particular DOAC over another in most patients, excepting that oral factor Xa inhibitors are preferred in patients with cancer [2,4].
Our preference for these agents is based upon their similar efficacy, reduced need for monitoring, and lower bleeding risk profile when compared with warfarin. Efficacy and safety across agents is similar. Dosing for each agent is per product information. When prescribing these agents, we prefer that they be administered by clinicians knowledgeable in their use in accordance with study criteria that proved their efficacy. (See 'Direct thrombin and factor Xa inhibitors' below and "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects" and "Management of bleeding in patients receiving direct oral anticoagulants" and 'Risk of bleeding' below.)
Patients in whom DOACs are not suitable include the following:
•Pregnant or lactating patients. (See "Venous thromboembolism in pregnancy and postpartum: Treatment" and "Anticoagulation during pregnancy and postpartum: Agent selection and dosing".)
•Patients with severe liver disease (defined by high Child-Pugh score B and C (table 5)). (See "Cirrhosis in adults: Overview of complications, general management, and prognosis", section on 'Child-Pugh classification'.)
•Patients with thrombotic antiphospholipid antibody syndrome, especially with a triple-positive antibody profile. (See "Antiphospholipid syndrome: Management".)
•Patients in whom monitoring compliance is important.
•Patients with hemodynamically unstable pulmonary embolism (PE) or massive iliofemoral deep vein thrombosis (DVT; eg, phlegmasia cerulea dolens). However, DOACs may be administered once the patient is stable and oral anticoagulation is being considered. (See "Venous thromboembolism: Initiation of anticoagulation".)
•Other patients – Management of DOACs in patients with kidney failure and obesity is discussed below. (See 'Direct thrombin and factor Xa inhibitors' below.)
●Warfarin – Warfarin is our preferred anticoagulant for patients in whom factor Xa or direct thrombin inhibitors are not available, for patients with antiphospholipid antibody syndrome, and for patients with severe kidney insufficiency in whom apixaban is not chosen (table 4). It is important to maintain good quality INR control with time in therapeutic range (TTR) >65 to 70 percent.
Warfarin management may not be feasible in patients with abnormal INR due to liver disease (LMW heparin may be preferred for treatment phase therapy in this population).
Warfarin requires multiple office visits for INR monitoring and has a higher bleeding risk when compared with DOACs; however, antidotes for warfarin-related bleeding are more readily available. Thus, it may also be preferred in those with a preference or requirement for a readily reversible antidote (eg, need for frequent interventions) and those in whom therapeutic anticoagulation needs to be monitored (eg, poor compliance). Dosing for warfarin is discussed below. (See 'Warfarin' below.)
Although warfarin is not absolutely contraindicated in patients with known protein C or S deficiency who are at increased risk of warfarin-induced skin necrosis, patients can be treated with warfarin overlapped with heparin, if DOACs cannot be administered. (See "Protein C deficiency", section on 'Warfarin-induced skin necrosis'.)
●LMW heparin – Subcutaneous LMW heparin is an acceptable alternative for nonpregnant patients in whom oral medications are not feasible (eg, malabsorption). It is the preferred therapy for DVT during pregnancy and is an alternative to oral factor Xa inhibitors in patients with active malignancy. Some experts also prefer this agent in those with liver disease since the elevated INR in this population may not reflect the effect of warfarin. Laboratory monitoring is not required. No agent is superior to the other and dosing for each agent is per product information. They are contraindicated in patients with severe kidney dysfunction (creatinine clearance <30 mL/minute); dosing adjustments for mild kidney insufficiency are listed in the table (table 6). (See 'Low molecular weight heparin' below.)
●Fondaparinux – Fondaparinux is a subcutaneous factor Xa inhibitor. Like LMW heparin, monitoring is not necessary in routine practice, and it is contraindicated in patients with severe kidney dysfunction. It is an alternative to LMW heparin when LMW heparin cannot be administered because of a history or diagnosis of heparin-induced thrombocytopenia (HIT). Fondaparinux should be used cautiously in patients with moderate kidney dysfunction (creatinine clearance <50 mL/minute) and is contraindicated in patients with severe kidney dysfunction (creatinine clearance <30 mL/minute). (See 'Fondaparinux' below and "Fondaparinux: Dosing and adverse effects".)
Special populations — Special populations of patients with acute VTE requiring specific consideration include those listed below (table 3).
Malignancy — The treatment of VTE and duration of anticoagulation in patients with malignancy is discussed in detail separately. (See "Anticoagulation therapy for venous thromboembolism (lower extremity venous thrombosis and pulmonary embolism) in adult patients with malignancy".)
Pregnancy — LMW heparin is the preferred agent for treatment phase (long-term) anticoagulation in pregnant individuals with acute VTE (table 7). The treatment of VTE, duration of therapy, and use of anticoagulants in pregnancy are discussed in detail separately. (See "Venous thromboembolism in pregnancy and postpartum: Treatment" and "Anticoagulation during pregnancy and postpartum: Agent selection and dosing".)
Heparin-induced thrombocytopenia — For patients with VTE and a diagnosis of HIT, anticoagulation with heparin, including UFH and LMW heparin, is contraindicated. Anticoagulation with a nonheparin anticoagulant should be administered. The management of patients with HIT is discussed in detail separately. (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia" and "Management of heparin-induced thrombocytopenia".)
Antiphospholipid antibody syndrome — For patients with antiphospholipid syndrome (APS), extended anticoagulation with warfarin is recommended, the details of which are discussed separately. (See "Antiphospholipid syndrome: Management".)
Transition from initiation to treatment phase — Full anticoagulation should be ensured during the transition from initial to treatment phase (long-term) therapy. Interruptions should be minimized during the first three months of anticoagulation because this is the period that has the highest risk of recurrent thrombosis. In some patients, the agent chosen for treatment phase use is the same agent that was selected for initiation phase (initial) anticoagulation (eg, low molecular weight [LMW] heparin, rivaroxaban, and apixaban), but in others, the initiation phase agent and the selected treatment phase agent belong to different classes, such that transitioning from one agent to another is necessary (eg, heparin to warfarin, heparin to edoxaban or dabigatran).
The optimal transition strategy varies with the treatment phase (long-term) anticoagulant chosen. There are four basic approaches to transitioning between these phases:
●Monotherapy (oral initiation) – Apixaban and rivaroxaban have higher doses used during the initiation phase (10 mg twice daily for seven days for apixaban, 15 mg twice daily for 21 days for rivaroxaban), which preclude the requirement for initial heparin therapy.
●Sequential therapy – Dabigatran and edoxaban do not have a higher dose studied for initiation, so a parenteral agent such as LMW heparin must be used for at least five days, after which the patient can be transitioned to dabigatran or edoxaban for the treatment phase. Medications are not given concomitantly (no overlap).
●Overlapping therapy – When warfarin is chosen for the treatment phase, it is generally started on the same day, following initial parenteral anticoagulant administration. Both agents are overlapped until at least five days of concomitant therapy have taken place, and the parenteral agent is not stopped until the INR has been ≥2 for at least 24 hours.
●Parenteral monotherapy – Certain patients (such as those unable to use oral agent, and selected patients with cancer-associated thrombosis) use LMW heparin for initiation phase therapy, and then continue this for the ongoing treatment phase. Depending on the specific LMW heparin used, the daily dose may be adjusted after the first 30 days of administration.
Transition strategy, dosing, and efficacy of specific anticoagulants are discussed in more detail in the sections below. (See 'Agents for treatment phase (long-term) anticoagulation' below.)
AGENTS FOR TREATMENT PHASE (LONG-TERM) ANTICOAGULATION
Direct thrombin and factor Xa inhibitors — For patients with acute VTE, factor Xa inhibitors (apixaban, edoxaban, rivaroxaban) and direct thrombin inhibitors (dabigatran) are first-choice oral anticoagulants for treatment phase anticoagulation in most nonpregnant patients who do not have APS or severe liver disease. Patients with severe kidney insufficiency or active cancer can use selected agents (table 4). (See 'Selection of agent' above.)
Direct thrombin and factor Xa inhibitors are fixed-dose oral agents that, unlike warfarin, do not require routine laboratory monitoring and dose adjustments. These agents reach their peak efficacy within one to four hours after ingestion. Unlike when transitioning to warfarin, in which a period of overlapping therapy between the initiation parenteral anticoagulant and warfarin is required, the DOAC is generally started when the continuous infusion of UFH is stopped or when the next dose of therapeutic LMW heparin would have been due (or slightly before). Specific recommendations for transitioning from parenteral anticoagulation to each DOAC are provided in the drug monographs included in UpToDate. Transitioning during anticoagulant therapy is discussed separately. (See 'Switching anticoagulants during therapy' below.)
Details regarding treatment of bleeding on these agents and reversibility agents are discussed separately. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects" and "Management of bleeding in patients receiving direct oral anticoagulants".)
Dosing — When an oral factor Xa or direct thrombin inhibitor is chosen for treatment phase anticoagulation, the initiation phase of therapy should be managed in accordance with study criteria that demonstrated its efficacy [10-15].
●Rivaroxaban and apixaban can be administered as monotherapy (ie, no overlap or prior treatment with parenteral anticoagulant [eg, LMW heparin] is necessary), but short courses of parenteral anticoagulant should be administered when there is a delay in obtaining these agents (eg, not available in local pharmacy).
●Edoxaban and dabigatran are administered following a five-day course of parenteral anticoagulation (not overlapped).
Typical initiation and treatment phase doses in those with normal kidney function are:
●Rivaroxaban 15 mg by mouth twice daily for 21 days followed by 20 mg once daily
●Apixaban 10 mg twice daily for seven days followed by 5 mg twice daily (2.5 mg twice daily for extended treatment beyond six months)
●Edoxaban 60 mg once daily (after five days of initiation phase therapy with a parenteral agent)
●Dabigatran 150 mg twice daily (after five days of initiation phase therapy with a parenteral agent)
Dosing may vary in specific populations:
●Kidney dysfunction – While most DOACs are avoided in those with severe kidney failure (eg, creatinine clearance [CrCl] <30 mL/minute as estimated by the Cockcroft-Gault equation), apixaban is labeled for use in all stages of kidney disease [16]. In our practice, for patients with VTE and an estimated CrCl <30 mL/minute, we favor initiating anticoagulation with unfractionated heparin for four to six days, then continuing therapy with apixaban 5 mg orally twice daily for the treatment phase in patients. Transition with overlapped warfarin is an alternative to apixaban. While apixaban is commonly used in patients with VTE who have severe kidney insufficiency (CrCl <30 mL/minute), there is a paucity of high-quality data to support it. Our approach in this population is based on ensuring therapeutic anticoagulation during initiation and avoiding a loading dose of 10 mg of apixaban, the efficacy of which is uncertain in severe kidney failure. Limited data in VTE patients suggest that apixaban at 5 mg twice daily provides sufficient therapeutic anticoagulation during the treatment phase [17,18]. Further data on DOAC dosing in kidney insufficiency are discussed separately. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Chronic kidney disease'.)
There is a boxed warning on the US label not to use edoxaban for patients with atrial fibrillation who have a creatinine clearance (CrCl) >95 mL/minute due to a numerically higher rate of ischemic stroke when compared with warfarin. This numerical increase in stroke with CrCl >95 mL/minute is also seen with other factor Xa inhibitor (apixaban, rivaroxaban) trials in patients with atrial fibrillation when compared with warfarin. This observation was not seen with the direct thrombin inhibitor dabigatran. Whether a similar increased risk of recurrent VTE occurs in this population is unknown. However, until these data are available in patients with VTE, we suggest avoiding the use of edoxaban in patients with a CrCl >95 mL/minute. When calculating CrCl, we suggest a more precise assessment if edoxaban is to be used (calculator 1). (See "Atrial fibrillation in adults: Use of oral anticoagulants", section on 'Dosing'.)
●Obesity – Other concerns regarding these agents include their distribution and efficacy in the obese population, which is understudied [19]. While randomized trials report efficacy for standard dosing in those with a body weight (BW) ≤100 kg or body mass index (BMI) ≤30 kg per m2, limited data from the International Society on Thrombosis and Hemostasis suggest similar efficacy in those with a BW >120 kg or BMI ≥40 kg per m2 [19], and small retrospective analyses suggested similar efficacy when compared with warfarin [20,21]. The efficacy is unclear in those with a BW between 101 to 119 kg or BMI 31 to 39 kg per m2. The ISTH suggest using rivaroxaban and apixaban at standard dosing rather than other DOACs in those with BW >120 kg or BMI ≥40 kg per m2.
●Others – Caution is advised when inducers or inhibitors of P-glycoprotein and cytochrome P450 are prescribed. Drug interactions with other agents are discussed in the table (table 8).
Manufacturer recommendations on compatibility of DOACs with feeding tube administration vary for each agent. For example, apixaban may be crushed and suspended in water whereas dabigatran capsules should not be opened due to substantial increase in drug exposure. These and other pharmacokinetic details are provided in the drug monographs included within UpToDate, "administration" field.
Efficacy — Large randomized trials and meta-analyses have reported the safety and efficacy of these agents for the treatment and prevention of recurrent VTE (deep vein thrombosis [DVT] and/or pulmonary embolism [PE]) [10-15,22-24]. Most of these trials were performed in stable patients and were designed as noninferiority trials that compare the newer agent with standard anticoagulation (ie, heparin followed by warfarin) and showed comparable safety and efficacy.
●Rivaroxaban – Rivaroxaban is an oral factor Xa inhibitor that demonstrated similar efficacy to conventional therapy (heparin followed by warfarin) for the treatment of acute VTE in a large, prospective, randomized controlled trial. For most patients receiving rivaroxaban, no parenteral anticoagulation was administered, and rivaroxaban was the initial anticoagulant used. EINSTEIN-DVT and EINSTEIN-PE were open-label randomized trials that enrolled a total of 8281 patients with acute DVT or PE and demonstrated the noninferiority of rivaroxaban (15 mg twice daily for 21 days followed by 20 mg once daily) to conventional therapy (enoxaparin followed by warfarin) for a treatment period of 3, 6, or 12 months [10,11]. Compared with conventional therapy, rivaroxaban resulted in similar rates of recurrent DVT (2.1 versus 3 percent), PE (2.1 versus 1.8 percent), and total bleeding events (8.1 versus 8.1 [EINSTEIN-DVT]; 10.3 versus 11.4 percent [EINSTEIN-PE]). In a post-hoc analysis of EINSTEIN-DVT and interrogation of a national registry, rivaroxaban was associated with a nonsignificant reduction in the rate of post-thrombotic syndrome [25,26]. (See "Overview of the treatment of lower extremity deep vein thrombosis (DVT)", section on 'Graduated compression stockings' and "Post-thrombotic (postphlebitic) syndrome in adults".)
●Apixaban – Apixaban is an oral factor Xa inhibitor. AMPLIFY was a prospective, randomized, double-blind trial that compared apixaban (10 mg twice daily for seven days for initial anticoagulation followed by 5 mg twice daily for six months) with conventional anticoagulation (subcutaneous enoxaparin for five days followed by warfarin for six months) in 5395 patients for the treatment of acute VTE (DVT and/or PE) [12]. There was no difference in the rates of recurrent symptomatic VTE or VTE-related death (2.3 versus 2.7 percent) between the groups, and fewer bleeding events were reported in the apixaban group (4.3 versus 9.7 percent). Subgroup analysis suggested that the efficacy of apixaban for the prevention of VTE or VTE-related deaths occurred in all patient groups (eg, DVT, PE, unprovoked VTE, extensive PE). Large-scale cohort analyses suggest similar efficacy in practice [27].
●Edoxaban – Edoxaban is an oral factor Xa inhibitor that has been shown to have a similar efficacy and superior safety profile when compared with warfarin for the treatment of acute VTE. In one trial, 4921 patients with acute VTE (DVT and/or PE) were randomized to receive 3 to 12 months of edoxaban or warfarin following initial therapy for five days with unfractionated or LMW heparin [13]. Edoxaban was administered orally at 60 mg once daily. A lower dose (30 mg once daily) was used for patients with a CrCl of 30 to 50 mL per minute or low body weight ≤60 kg. Compared with warfarin, edoxaban had a similar rate of recurrent symptomatic VTE or VTE-related death (3.2 versus 3.5 percent) and fewer bleeding events (8.5 versus 10.3 percent). A post-hoc analysis of the patients on low dose edoxaban showed similar results [28].
Of potential clinical interest was the superior efficacy in the prespecified subgroup of patients with PE that had right ventricular (RV) dysfunction as assessed by elevated brain (or beta) natriuretic peptide (BNP) or increased RV dimensions on computed tomography (CT). However, BNP is a nonspecific biomarker of RV dysfunction, and echocardiography is a more accurate test than CT for the assessment of RV dysfunction. Further study of this agent in this population of patients with severe PE and RV dysfunction by echocardiogram is warranted to validate these findings.
●Dabigatran – Dabigatran is an oral direct thrombin inhibitor. A large, randomized controlled trial suggested that dabigatran has similar efficacy to warfarin for the prevention of recurrent VTE. However, concerns have been raised about its efficacy and risk of thrombosis given the broad inferiority margin set in one of the trials (RE-COVER). In a randomized, double-blind trial (RE-COVER I), 2539 patients with acute VTE were treated for six months with either dabigatran (150 mg by mouth twice per day) or warfarin, each after seven days of initial parenteral anticoagulation [14]. Compared with warfarin, dabigatran had a similar incidence of recurrent VTE or VTE-related deaths (2.4 versus 2.1 percent), VTE-related deaths (0.1 versus 0.2 percent), major bleeding events (1.6 versus 1.9 percent), and any bleeding event (16.1 versus 21.9 percent). These results suggest that the efficacy and safety profile of dabigatran is similar to that of warfarin for the treatment of acute VTE. Similar results were reported in an identically designed trial of 2589 patients with acute DVT (RECOVER II) [15] and in a pooled analysis of RE-COVER I and II [29]. In a post hoc analysis of these trials, the efficacy appears to be maintained in patients with thrombophilia and antiphospholipid syndrome [30].
Meta-analyses support individual findings. As an example, one meta-analysis of 64 studies that compared DOACs with standard anticoagulation, usually warfarin, reported no difference in VTE recurrence rate (risk reduction [RR] 0.93, 95% CI 0.53-1.63 [oral direct thrombin inhibitors]; RR 0.85, 95% CI 0.63-1.13 [factor Xa inhibitors]) and possibly lower bleeding rates with DOACs (RR 0.51, 95% CI 0.15-1.67 [oral direct thrombin inhibitors]; RR 0.91, 95% CI 0.56-1.48 [factor Xa inhibitors]) [4].
No randomized trials have directly compared individual DOACs with each other. However, retrospective data suggest that apixaban may have lower rates of recurrent VTE and bleeding than rivaroxaban [31-33]. As an example, in data derived from an insurance database, among 37,000 new users of apixaban and rivaroxaban, apixaban was associated with a lower rate of recurrent VTE (hazard ratio [HR] 0.77, 95% CI 0.69-0.87) and bleeding (HR 0.60, CI 0.53-0.69) [33]. Reasons for this difference are unknown but could be due to differences in the population characteristics or differences in the pharmacokinetics (eg, less fluctuations in drug concentration with apixaban [34]).
Retrospective data suggest that efficacy of DOACs compared with warfarin is maintained in older patients (eg, >80 years) [35].
The bleeding risk of these agents is discussed in greater detail below. (See 'Oral factor Xa and direct thrombin inhibitors' below.)
Warfarin
Dosing — Warfarin is the preferred agent for nonpregnant patients in whom factor Xa or direct thrombin inhibitors are not available and for patients with APS. In patients with severe kidney dysfunction, warfarin is a preferred agent (though apixaban may also be an option, depending on the indication) [17,18]. It may also be preferred in those in whom therapeutic anticoagulation needs to be closely monitored (table 4). (See 'Selection of agent' above.)
When warfarin is chosen as the agent for treatment phase anticoagulation, it is generally started on the same day with LMW heparin or UFH (day 1). Several approaches to initiation are available, including use of clinical nomograms [36,37] (which can be modified with pharmacogenetic information, though the benefit of this information is uncertain [38,39]) or at a typical initiating dose of 5 mg/day for the first two days (range 2 to 10 mg/day) (table 9) [40,41]. Initial doses at the lower range (2 to 5 mg/day) may be considered in those assessed at high bleeding risk (eg, older adults), and doses in the higher range (5 to 10 mg/day) may be selected in healthy individuals who are at low risk for bleeding. Regardless of the method used for starting dose determination, dosing is then adjusted until the INR is within the therapeutic range (2 to 3; target 2.5) for two consecutive days. INR monitoring is continued, generally with decreasing frequency, as stable dosing is achieved. Adequate INR control (as reflected by time in therapeutic range [TTR], ideally ≥70 percent) is a major determinant of effectiveness and safety in patients with atrial fibrillation, and this also likely applies to patients with VTE. (See "Warfarin and other VKAs: Dosing and adverse effects", section on 'Importance of strict INR control' and "Venous thromboembolism: Initiation of anticoagulation", section on 'Anticoagulant agents'.)
Warfarin cannot be administered as the only initial anticoagulant for the treatment of patients with VTE. When chosen as the treatment phase (long-term) anticoagulant it must be coadministered with heparin so that full anticoagulation is assured. The premature cessation of heparin before warfarin has taken its full effect (ie, therapeutic INR for two days) may result in inadequate protection against recurrent thrombosis. This overlap is required because it takes a number of days for all of the vitamin K-dependent factors to become depleted (factors II, VII, IX, and X). During the first few days of warfarin therapy, prolongation of the prothrombin time (PT; as reflected by the INR) mainly reflects the depression of factor VII, which has a short half-life (hours) (figure 2). Importantly, this does not represent adequate anticoagulation because additional vitamin K-dependent factors are insufficiently reduced (takes approximately five days with adequate dosing). In addition, reduction of protein C and S occurs shortly after warfarin therapy and potentially renders a procoagulant state. Together, these factors increase the likelihood of recurrent thrombosis should patients not be fully anticoagulated with parenteral heparin. (See "Overview of hemostasis" and "Warfarin and other VKAs: Dosing and adverse effects", section on 'Warfarin administration'.)
Efficacy — Data that support the efficacy of warfarin as an anticoagulant in patients with VTE are derived from older studies that compared warfarin with no anticoagulation or low-dose subcutaneous heparin and from randomized trials and meta-analyses of variable durations of therapeutic anticoagulation [42-52]. In the seminal trial, performed in 1960, that compared warfarin with observation in patients with acute DVT, warfarin resulted in a dramatic reduction in recurrence (0 versus 26 percent), which translated into a mortality benefit (26 versus 0 percent) [52]. Since then, most other trials have compared various durations of warfarin therapy to provide an estimate of the risk of recurrence [43,51]. Cumulatively, these data all support low rates of recurrent VTE and death in patients treated with warfarin therapy for proximal DVT, with the greatest benefit occurring within the first few days or weeks of the initial event. As an example, one 2010 meta-analysis of 13 prospective cohort studies and 56 randomized clinical trials reported rates of recurrent VTE and fatal VTE during the first three months of warfarin therapy as 3.4 and 0.4 percent, respectively [51].
Low molecular weight heparin
Dosing — Low molecular weight (LMW) heparin is the preferred agent for those in whom treatment with one of the oral agents is not feasible (eg, patients with poor or no oral intake). (See 'Selection of agent' above.)
When subcutaneous LMW heparin is chosen for treatment phase anticoagulation and UFH is the initial anticoagulant being used, LMW heparin can be administered and the UFH infusion immediately discontinued.
The dosing of LMW heparin is individualized according to each product and described separately. (See "Venous thromboembolism: Initiation of anticoagulation", section on 'Low molecular weight heparin' and "Heparin and LMW heparin: Dosing and adverse effects", section on 'LMW heparin'.)
Efficacy — Several randomized trials and meta-analyses have shown that LMW heparin is at least as effective as warfarin in the prevention of recurrent VTE (DVT and/or PE) with a similar rate of major bleeding and mortality [53-55]. However, many of the trials have potential bias for the assessment of recurrent VTE and imprecision in the assessment of major bleeding and mortality, which limits the analysis.
As examples:
●A 2017 meta-analysis of 16 trials of patients with VTE (total 3299 patients) compared LMW heparin with warfarin [54]. LMW heparin was as effective as warfarin in preventing symptomatic VTE (odds ratio [OR] 0.83; 95% CI 0.6-1.15) and was associated with a nonsignificant trend in lower bleeding rates (OR 0.62, 95% CI 0.36-1.07). No difference in mortality was observed.
●Another 2012 meta-analysis of eight trials that restricted analysis to only those trials that used ≥50 percent of the full therapeutic dose of LMW heparin, reported a reduction in recurrent VTE compared to warfarin (RR 0.62, 95% CI 0.46-0.84) with a similar rate of major bleeding and mortality [53].
Randomized trials have also suggested that, compared with warfarin, treatment phase (long-term) anticoagulation with LMW heparin lowers the frequency of post-thrombotic syndrome (PTS), a treatable, late complication of acute DVT [56-58]. One 2011 systematic review of nine trials of patients with DVT reported that, compared with other agents (mostly warfarin), LMW heparin was associated with a reduction in the rate of PTS signs and symptoms (OR 0.77), recanalization of thrombosed veins (RR 0.66), and venous ulceration (RR 0.13) [58]. However, the major indication for anticoagulation is the prevention of recurrence rather than the prevention of PTS such that the latter does not practically influence agent selection. (See "Post-thrombotic (postphlebitic) syndrome in adults" and "Overview of the treatment of lower extremity deep vein thrombosis (DVT)", section on 'Graduated compression stockings'.)
Treatment phase anticoagulation with LMW heparin has never been directly compared with factor Xa or direct thrombin inhibitors. (See 'Direct thrombin and factor Xa inhibitors' above.)
Fondaparinux — Fondaparinux, a pentasaccharide, is an alternative to LMW heparin when the latter is contraindicated in patients with heparin-induced thrombocytopenia (HIT). (See 'Selection of agent' above.)
When transitioning from initial anticoagulation with UFH, subcutaneous fondaparinux can be administered and the UFH infusion immediately discontinued. Fondaparinux is typically dosed according to the usual initiating dose. Subcutaneous fondaparinux is contraindicated in patients with severe kidney insufficiency (CrCl <30 mL/minute) and may accumulate in patients with moderate kidney insufficiency (CrCl 30 to 50 mL/minute). Information on dosing is provided separately. (See "Fondaparinux: Dosing and adverse effects".)
Although fondaparinux is less well studied than LMW heparin as a treatment phase anticoagulant, fondaparinux appears to have a similar efficacy and safety profile to the LMW heparin, enoxaparin when used as an initiation phase anticoagulant. A meta-analysis of pentasaccharides that included fondaparinux also reported similar efficacy when fondaparinux was given in combination with warfarin [59].
Detailed discussion of fondaparinux dosing and adverse effects are provided separately. (See "Fondaparinux: Dosing and adverse effects".)
Unfractionated heparin (subcutaneous) — Subcutaneous UFH may be an alternative in patients who are unable or unwilling to take warfarin or LMW heparin (eg, patients with kidney insufficiency unable to take an oral medication). In one randomized open-label study of 697 patients with acute DVT, there were no differences in efficacy or bleeding rates between fixed-dose subcutaneous UFH (333 units/kg initially, followed by 250 units/kg twice daily) and low molecular weight heparin (enoxaparin/dalteparin) when used as initial anticoagulants and overlapped with warfarin [60]. However, treatment phase anticoagulation with subcutaneous UFH has been inadequately studied and has never been compared with factor Xa or direct thrombin inhibitors.
DURATION OF TREATMENT PHASE AND TRANSITION TO EXTENDED PHASE
Most patients — The typical duration of the treatment phase of anticoagulation is at least three months, rather than a shorter treatment duration of four to six weeks [2,4]. Selected patients also require extended (indefinite) anticoagulation, as below. (See 'Candidates for extended phase therapy' below.)
Evidence supporting anticoagulation for a minimum of three months is derived from randomized trials and meta-analyses of patients with acute VTE, the majority of whom had an unprovoked event, and proximal DVT and/or PE:
●Without therapeutic anticoagulation, the risk of recurrent venous VTE (DVT and PE) is highest during the first three months following the initial event [42-50,53].
●Shortening the course of anticoagulation from three or six months to four or six weeks is associated with an increased risk for recurrent VTE, even in patients with distal DVT [44,48,49,53,61]. As an example, in a 2012 meta-analysis of five randomized trials, four to six weeks of warfarin therapy resulted in an increase in the absolute risk of recurrent VTE compared with three months of therapy (53 more episodes of VTE per 1000), and only a small decrease in the risk of major bleeding (five fewer major bleeding episodes per 1000) [53].
Although patients with distal DVT have a lower risk of recurrence (approximately 1 percent per year), we typically treat distal DVT for a duration of three months based upon the following data [4,42,61,62]:
●In a meta-analysis of randomized trials and cohort studies of patients with isolated distal DVT, a lower rate of recurrent VTE was observed in patients anticoagulated for more than six weeks compared with patients treated with a six-week course of anticoagulant therapy (odds ratio [OR] 0.50, 95% CI 0.31-0.79) [62].
●A subsequent randomized trial of patients with symptomatic distal DVT showed reduced rates of recurrent DVT without an increase in bleeding in patients treated with three months of therapy as compared with six weeks [61]. Therefore, based on these data and extrapolation from patients with proximal DVT, we prefer that patients with distal DVT be anticoagulated for three months [4,42].
Data that support a three-month course of anticoagulation in patients with asymptomatic, incidental, or small subsegmental PE are discussed separately. (See "Acute pulmonary embolism in adults: Treatment overview and prognosis", section on 'Patients with subsegmental PE'.)
Candidates for longer finite (6 to 12 month) treatment — In selected populations, the treatment phase may be longer. Examples include:
●Pregnancy-associated VTE – Treatment phase anticoagulation is not stopped until at least three months of therapy and not until six weeks postpartum. (See "Venous thromboembolism in pregnancy and postpartum: Treatment", section on 'Duration of therapy'.)
●Cancer-associated thrombosis – Some patients with cancer-associated thrombosis are treated with a six-month treatment phase, followed by transition to extended duration (indefinite) anticoagulation in most patients, while the cancer is active. (See "Anticoagulation therapy for venous thromboembolism (lower extremity venous thrombosis and pulmonary embolism) in adult patients with malignancy", section on 'Duration of anticoagulation and follow-up'.)
●Thrombosis at unusual sites – In patients with splanchnic or cerebral veins, treatment phase lengths of 6 to 12 months have been commonly used. (See "Mesenteric venous thrombosis in adults", section on 'Anticoagulation' and "Cerebral venous thrombosis: Treatment and prognosis", section on 'Duration of anticoagulation'.)
●Extensive or life-threatening episodes of provoked VTE – Expert opinion supports treatment phase anticoagulation for longer finite periods (or indefinite anticoagulation) in patients with extensive or life-threatening disease [63]. The rationale is that recurrent episodes may be more likely to be extensive or life-threatening in these patients.
●Patients with provoked VTE and persistent reversible risk factors – Patients with provoked VTE and persistent reversible risk factors (eg, prolonged immobility, continued estrogen use) have higher risk of VTE recurrence. The presence of the provocation, until resolved, likely increases the risk of VTE recurrence. If a risk factor persists over time but will eventually be mitigated, the treatment phase is generally extended until the risk factor can be addressed [42,49,53].
In general, randomized trials and meta-analyses that mostly include patients at low-risk of recurrence do not definitively show a reduction in VTE recurrence with a more prolonged, finite period of treatment beyond three months (eg, 6 or 12 months) [49,53,64]. As an example, in a 2025 meta-analysis of 22 trials, a longer treatment course of >6 months, compared with treating for three to six months, may reduce the risk of recurrent PE (risk ratio [RR] 0.66, 95% CI 0.42-1.02) and DVT (RR 0.85, 95% CI 0.63-1.14) but at the expense of higher risk for major bleeding (RR 2.02, 95% CI, 1.02-3.98) [64]. Prior meta-analyses did not demonstrate convincing reduction in recurrence but did show an increase in bleeding [49,53].
However, the rationale for increasing the duration of the treatment phase in higher-VTE recurrence risk patients is that many trials excluded such patients. This disparity between clinical practice and published trials is due to trial exclusion of patients that may theoretically benefit from longer courses of anticoagulation (eg, extensive or life-threatening VTE).
In addition, some data suggest that the risk of recurrence does continue to decline between three and six months on anticoagulant therapy [53,64]. For these reasons, some guidelines and many experts use a three- to six-month range for an optimal duration of therapy for selected patients with provoked VTE who are assessed to have a higher VTE recurrence risk and have low risk of bleeding [42]. On the other hand, completing therapy at three months is also appropriate for patients with a moderate to high bleeding risk.
While practice varies, we do not obtain a repeat ultrasound at the end of the treatment phase to guide a flexible duration of treatment. One nonblinded trial showed that an ultrasound-directed flexible duration of treatment reduced recurrent VTE rates compared with a fixed duration of three months (12 versus 17 percent) [65]. However, it is possible that ultrasound-directed treatment may not identify patients at higher VTE risk but rather extend anticoagulation and prevent recurrence in patients for whom a longer duration would have been provided regardless.
Candidates for extended phase therapy — Regardless of the chosen duration of the treatment phase of therapy, at its conclusion, a decision must be made whether a patient should be offered extended phase (indefinite) therapy with the goal of secondary prevention of recurrent VTE. It is critical that this decision be individualized according to the presence or absence of provoking events and risk factors, risk for recurrence and bleeding (table 10 and table 1), as well as to the individual patient's preferences and values. Our approach is consistent with guidelines set out by the American College of Chest Physicians, the International Society of Thrombosis and Hemostasis, and the International Consensus Statement on the Prevention and Treatment of Venous Thromboembolism [4,42,53,66].
Selecting patients for extended phase anticoagulation is discussed separately. (See "Selecting adult patients with lower extremity deep venous thrombosis and pulmonary embolism for indefinite anticoagulation" and "Anticoagulation therapy for venous thromboembolism (lower extremity venous thrombosis and pulmonary embolism) in adult patients with malignancy", section on 'Duration of anticoagulation and follow-up'.)
SWITCHING ANTICOAGULANTS DURING THERAPY —
Interruptions should be limited especially during the first three months of anticoagulation. However, anticoagulants may need to be changed for medical reasons as well as for altered patient preferences. Reasons for switching agents include:
●The development of kidney insufficiency (prolongs the half-life of low molecular weight [LMW] heparin, fondaparinux, and factor Xa and direct thrombin inhibitors [ie, direct oral anticoagulants (DOACs)])
●Perceived burdens of laboratory testing for warfarin
●Poor compliance or difficulty with international normalized ratio (INR) testing
●Resolution of active cancer
●Pain or inflammation at injection sites
●Cost or changes in medication coverage
●Need for repeated invasive procedures
●Recurrence despite therapeutic anticoagulation
Clinicians should alert their patients that switching anticoagulants is associated with a potential increase in the risk of both bleeding and recurrence. These risks are less well studied in patients transitioning to and from factor Xa and direct thrombin inhibitors [67]. When switching to or from these agents, we prefer to use protocols that are similar to those used in trials that studied these agents in patients with atrial fibrillation and VTE.
In general, the following applies:
●Transitioning from LMW heparin:
•To warfarin should be the same as for initial anticoagulation. Warfarin and heparin are simultaneously administered for four to five days until the INR is therapeutic for a minimum of 24 hours or two consecutive days. (See 'Agents for treatment phase (long-term) anticoagulation' above and "Warfarin and other VKAs: Dosing and adverse effects", section on 'Warfarin administration' and "Venous thromboembolism: Initiation of anticoagulation", section on 'Duration of therapy for heparin'.)
•To oral factor Xa inhibitor or oral thrombin inhibitor is generally done by administering the oral agent within zero to two hours before the next scheduled dose of LMW heparin is due. Specific recommendations for transitioning from parenteral anticoagulation to each DOAC are provided in the drug monographs included in UpToDate for each DOAC. (See 'Direct thrombin and factor Xa inhibitors' above.)
●Transitioning from unfractionated heparin:
•To warfarin is discussed separately; principles are similar to LMW heparin. (See 'Warfarin' above.)
•To a DOAC is usually done by starting the DOAC when a continuous infusion of UFH is stopped as DOACs have a rapid onset of action, with a peak anticoagulant effect occurring two to three hours after intake. Specific recommendations for transitioning from parenteral anticoagulation to each DOAC are provided in the drug monographs included in UpToDate for each DOAC.
•Direction on switching oral agents is provided separately (table 11). The biggest challenge is switching from a DOAC to warfarin. This challenge can be averted by stopping the DOAC and beginning LMW heparin overlapping warfarin according to the standards described above. (See 'Warfarin' above and "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Transitioning between anticoagulants'.)
Perioperative management of anticoagulation is discussed separately. (See "Perioperative management of patients receiving anticoagulants".)
MONITORING —
All patients on anticoagulation should be monitored clinically for therapeutic efficacy (recurrence), bleeding, as well as the development of conditions that affect the half-life of the medications used (eg, kidney failure, pregnancy, weight gain/loss) and adverse effects of the medications (eg, skin necrosis, thrombocytopenia, osteoporosis). Laboratory monitoring varies with the long-term anticoagulant used:
●Oral factor Xa and direct thrombin inhibitors – These agents do not require routine laboratory monitoring because they are administered in fixed-dose regimens without dose adjustments and laboratory parameters have not been correlated with clinical endpoints. They should be administered by physicians knowledgeable in their use and in accordance with study criteria that proved their efficacy. Patients should be carefully monitored clinically for bleeding and for kidney failure. Monitoring therapeutic efficacy is difficult but is performed on rare occasions, for example in patients at the extremes of body weight (ie, <45 kg or >130 kg). In addition, a practice tool aimed at monitoring patients for the complications of these agents is published but not routinely used [68]. Further details regarding monitoring patients on these agents are discussed separately. (See 'Direct thrombin and factor Xa inhibitors' above and "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects".)
●Warfarin – The most common laboratory test used to monitor warfarin is the prothrombin time (PT) ratio, usually expressed as the international normalized ratio (INR). The goal INR is 2 to 3 (target 2.5) [41,69]. Once the anticoagulant effect and the patient's warfarin dose requirements are stable, the INR should be monitored usually every three to four weeks throughout the course of warfarin therapy for VTE. More frequent monitoring is indicated if factors are present that may produce an unpredictable response to warfarin (eg, concomitant therapy with drugs that interact with warfarin); longer intervals (eg, up to 12 weeks) are feasible if the INR response is predictable [41]. Warfarin dosing must be individualized because factors including drug interactions, increased age, and specific genotypes alter the response to warfarin [70]. Genotype-guided dosing has been studied as a means of selecting a starting dose of warfarin; and is of questionable benefit. It has no proven value over conventional monitoring in ongoing management. Self-monitoring (obtaining an INR using a home device and contacting a health professional for dose adjustment) and self-management (measuring INR and adjusting warfarin dose independently) have proven efficacy in well-selected patients who demonstrate competency in these strategies. Further details regarding monitoring patients on warfarin are discussed separately. (See "Warfarin and other VKAs: Dosing and adverse effects", section on 'Warfarin administration' and "Warfarin and other VKAs: Dosing and adverse effects", section on 'Self-monitoring and self-management' and "Clinical use of coagulation tests", section on 'Prothrombin time (PT) and INR'.)
●Low molecular weight heparin and fondaparinux – These agents do not require routine laboratory monitoring because laboratory parameters have not been correlated with clinical endpoints (recurrence and bleeding). However, patients should be monitored for the development of kidney insufficiency, which can prolong their half-life. If there is any question as to the correct dose of low molecular weight (LMW) heparin (eg, for those with obesity or kidney insufficiency), measuring antifactor Xa activity testing is sometimes used. Fondaparinux was not monitored in clinical studies and is therefore not monitored routinely in clinical practice. Measurement of fondaparinux drug levels may be appropriate in patients with major bleeding, although fondaparinux-calibrated anti-Xa activity assays are not widely available. Details regarding the therapeutic use of LMW heparin and fondaparinux and issues that may affect interinstitutional variability among tests used to monitor these agents are discussed separately. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'LMW heparin' and "Fondaparinux: Dosing and adverse effects", section on 'Pharmacology' and "Clinical use of coagulation tests", section on 'Monitoring heparins'.)
There are no data to support the use of incentives or alerts to improve adherence to anticoagulants.
RISK OF BLEEDING —
All patients should be regularly assessed during anticoagulant therapy for bleeding risk (table 12). When reversal of anticoagulation is needed as part of a bleeding management strategy, the reversal strategy differs by agent. (See "Management of bleeding in patients receiving direct oral anticoagulants" and "Heparin and LMW heparin: Dosing and adverse effects", section on 'Reversal' and "Management of warfarin-associated bleeding or supratherapeutic INR".)
Warfarin and low molecular weight heparin — Evidence from randomized trials and meta-analyses suggest that the rate of major bleeding on warfarin and low molecular weight (LMW) heparin are generally low. As an example, in a meta-analysis of 13 prospective cohort studies and 56 randomized clinical trials of patients treated for deep vein thrombosis (DVT), rates of major bleeding and major fatal bleeding were 1.6 and 0.2 percent, respectively, during the first three months of anticoagulant therapy [51].
Rates of any bleeding events are usually higher, ranging from 5 to 15 percent, and the risk of bleeding is almost twice as high with the concomitant use of additional agents including nonsteroidal anti-inflammatory drugs (NSAIDs) or aspirin (hazard ratio 2.37 and 1.5, respectively) [71].
However, reported rates may not accurately reflect those expected to occur in clinical practice, where the risk of bleeding varies significantly among individuals.
Oral factor Xa and direct thrombin inhibitors — For patients on factor Xa and direct thrombin inhibitors, rates of major bleeding events range from 0.6 to 1.6 percent, and any bleeding events range from 4 to 16 percent, in meta-analyses of randomized trials [23,24].
Rates of bleeding are higher with concomitant use of NSAIDs (hazard ratio [HR] for hospital-diagnosed bleeding 2.09, 95% CI 1.67–2.62) or antiplatelet agents (odds ratio [OR] for major bleeding 1.89, 95% CI 1.04‐3.44) [72,73].
In randomized trials and meta-analyses of patients with VTE, rates of bleeding are lower with direct oral anticoagulants compared with warfarin [22-24,74-81]. As an example, in two 2023 meta-analyses of randomized trials, including all of the major studies discussed above, use of direct thrombin or factor Xa inhibitors resulted in reduced rates of bleeding when compared with conventional anticoagulation (heparin plus warfarin) (OR 0.63, 95% CI 0.45-0.89 for DVT; OR 0.71, 95% CI 0.36-1.41 for pulmonary embolism) [23,24].
Although bleeding rates may be lower with oral factor Xa and direct thrombin inhibitors, it is important to note that reported rates reflect those of study patients with a baseline low bleeding risk and may not represent those of the general population.
●Real-world bleeding risk with dabigatran – As an example, in the months following the approval of dabigatran in other settings, the US Food and Drug Administration (FDA) received many reports of serious and fatal bleeding events associated with DOAC use, and reports of bleeding rates in older populations have been conflicting [15,82-86]:
•An independent analysis of bleeding events and bleeding-related fatalities from reports submitted to the FDA (from 2011 through 2012) showed that higher numbers of bleeding events and fatal bleeding occurred with dabigatran compared with warfarin. However, the precise magnitude of difference is difficult to ascertain from the data source, which consisted of voluntary reports of adverse events [84]. This could be attributed to reporting bias after drug approval (a so-called “Weber effect”) [87].
•While a 2013 mini-Sentinel analysis by the FDA suggested that dabigatran was being used in accordance with its labeled indications and bleeding rates did not appear to be higher than those associated with warfarin, a subsequent 2014 FDA analysis reported an increased rate of gastrointestinal bleeding (GIB) among patients older than 65 years [82,83].
•A pooled analysis of bleeding events with dabigatran in RECOVER I/II reported that a lower bleeding risk with dabigatran was only evident in younger patients and that, in those over the age of 85 years, the bleeding risk was lower with warfarin [15].
•The Institute for Safe Medication Practices analyzed 1734 reports to the FDA from patients on warfarin, dabigatran, and rivaroxaban [85]. Dabigatran bleeds were five times more likely than warfarin to result in death (19 versus 4 percent; OR 5.2, 95% CI 3.4-8). The estimate of higher odds for a fatal outcome with dabigatran persisted after adjusting for age, sex, and report source.
•Postmarketing reports of bleeding, particularly for dabigatran, dampened initial enthusiasm for its use as anticoagulant therapy for VTE [82,83,88]. Increased risk of bleeding has also been reported in older patients (>75 years) on dabigatran [15,89,90]. Dabigatran is renally excreted, and, although unproven, this risk may be due to the higher prevalence of kidney insufficiency and longer half-life of dabigatran in older patients [91].
●Other reports on the safety of oral factor Xa and direct thrombin inhibitors – Other reports reflective of their use in general practice are limited.
•One 2016 retrospective analysis reported that 80 percent of patients with VTE were treated with warfarin and that warfarin and direct thrombin or factor Xa inhibitors were associated with similar rates of recurrent VTE (1.2 versus 2.1 percent) and bleeding (0.5 percent) [92].
•Another retrospective review reported that over one-half of patients prescribed direct thrombin or factor Xa inhibitors (for atrial fibrillation and VTE) were underdosed, when compared with the manufacturer recommendations [93]. The rate of thromboembolic events was 11 percent (apixaban), 4 percent (rivaroxaban), and 5 percent (dabigatran); bleeding event rates were 18 percent (apixaban), 18 percent (rivaroxaban), and 24 percent (dabigatran). Some of the factors that may have influenced reduced-dose prescribing were previous history of bleeding and potential drug interactions.
●Bleeding risk with rivaroxaban – Among the DOACs, rates of GIB are likely higher with rivaroxaban than other DOACS. One 2018 retrospective analysis of anticoagulant treatment (mostly given for atrial fibrillation) reported higher rates of upper GIB with rivaroxaban, while the lowest rates were in patients receiving apixaban [94-96]; patients who were also receiving a proton pump inhibitor experienced reduced bleeding from all anticoagulants. A 2021 analysis that included over 3000 patients also reported that rivaroxaban had the highest rates of GIB when compared with apixaban (all GIB events: 3.2 versus 2.5 events per 100 person-years; major GIB events: 1.9 versus 1.4 events per 100 person-years) or dabigatran (all GIB events: 3.2 versus 1.9 events per 100 person-years; major GIB: 1.9 versus 1.4 events per 100 person-years) [95].
Retrospective data suggests that bleeding rates among octogenarians receiving a DOAC appear to be at least equivalent (or lower) compared to warfarin [35,97].
The treatment of bleeding associated with DOACs is discussed separately. (See "Management of bleeding in patients receiving direct oral anticoagulants".)
Risk factors and predictive tools — Risk factors for bleeding on anticoagulation therapy in patients with VTE are presented elsewhere. (See "Overview of the treatment of lower extremity deep vein thrombosis (DVT)", section on 'Pretreatment assessment of bleeding risk' and "Risks and prevention of bleeding with oral anticoagulants".)
Although tools are available for estimating the risk of bleeding on anticoagulation (eg, HAS-BLED score (calculator 2) [98-100] and VTE-BLEED have been both derived and validated in the VTE population [101,102]), their predictive value in a particular patient is modest. However, such clinical bleeding risk scores should be used to draw attention to modifiable bleeding risks, and to identify the high bleeding risk patients for early review and follow-up.
Scoring systems that estimate the risk of bleeding are discussed separately. Newer scores have been derived and have undergone limited validation but await more data before being ready for widespread clinical adoption [103]. (See "Selecting adult patients with lower extremity deep venous thrombosis and pulmonary embolism for indefinite anticoagulation", section on 'Bleeding risk models'.)
While abnormal uterine bleeding (AUB) is not a contraindication per se to anticoagulation, retrospective data suggest that anticoagulation can increase the rate of AUB and have a negative impact on the quality of life [104].
RECURRENT VENOUS THROMBOEMBOLISM ON ANTICOAGULATION —
Some patients may have a thromboembolic event while receiving anticoagulation (sometimes called "event-on-therapy thrombosis"). In these patients, the diagnosis of recurrence should be confirmed with radiologic testing (eg, repeat compression ultrasonography or CT pulmonary angiography), ideally compared directly with previous imaging. Subtherapeutic anticoagulation is the most common reason, but several other etiologies including ongoing thrombotic stimulus (eg, cancer, antiphospholipid syndrome) may be present. Occasionally, no obvious cause is evident (idiopathic). The causes and management of suspected recurrence are discussed separately. (See "Acute pulmonary embolism in adults: Treatment overview and prognosis", section on 'Management of recurrence on therapy' and "Anticoagulation therapy for venous thromboembolism (lower extremity venous thrombosis and pulmonary embolism) in adult patients with malignancy", section on 'Management of recurrence'.)
SOCIETY GUIDELINE LINKS —
Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Superficial vein thrombosis, deep vein thrombosis, and pulmonary embolism" and "Society guideline links: Anticoagulation".)
INFORMATION FOR PATIENTS —
UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topics (see "Patient education: Deep vein thrombosis (blood clot in the leg) (The Basics)")
●Beyond the Basics topics (see "Patient education: Deep vein thrombosis (DVT) (Beyond the Basics)" and "Patient education: Warfarin (Beyond the Basics)")
SUMMARY AND RECOMMENDATIONS
●Terminology – Treatment phase (long-term) anticoagulant therapy for venous thromboembolism (VTE) is administered beyond the initiation phase of anticoagulation for a finite period (usually 3 but up to 12 months). (See 'Terminology' above.)
●Indications – Options for treatment phase anticoagulation include oral anticoagulants (ie, factor Xa inhibitors, direct thrombin inhibitors, and warfarin) and parenteral subcutaneous anticoagulants (low molecular weight [LMW] heparin and fondaparinux). While the oral factor Xa and direct thrombin inhibitors are typically preferred, choosing among these options frequently depends upon clinician experience and availability, the risks of bleeding, patient comorbidities and preferences, cost, and convenience (table 4 and table 3). In general, the following applies (see 'Selection of agent' above):
•First-line agents – For most nonpregnant patients, we recommend a direct oral anticoagulant (DOAC; ie, apixaban, edoxaban, rivaroxaban, or dabigatran) rather than other agents (Grade 1B). In general, these agents have similar efficacy to warfarin and a lower risk of bleeding.
DOACs should not be used in patients with thrombotic antiphospholipid syndrome, pregnant or lactating individuals, patients with severe liver disease, patients with severe kidney insufficiency (although apixaban is the exception), patients with extremes of weight (<45 or >130 kg), and patients in whom monitoring compliance is important. (See 'Selection of agent' above and 'Direct thrombin and factor Xa inhibitors' above.)
•Second-line agents
-Warfarin is an alternative for patients in whom there is concern about the poor availability of reversal agents and in those with severe kidney insufficiency (although apixaban may be used in those with a CrCl <15 mL/minute). Warfarin should be overlapped with heparin until the international normalized ratio is therapeutic for 24 hours. (See 'Selection of agent' above and 'Warfarin' above and 'Monitoring' above.)
-LMW heparin and fondaparinux are also effective treatments and may be preferred in some special populations without severe kidney failure. Dosing requirements are product specific, and no laboratory monitoring is required. (See 'Selection of agent' above and 'Low molecular weight heparin' above and 'Fondaparinux' above and 'Monitoring' above.)
-Anticoagulant therapy in patients who are pregnant and in selected patients who have active malignancy, in whom LMW heparin is the long-term agent of choice, and patients with heparin-induced thrombocytopenia, in whom heparin should be avoided, are discussed separately. (See 'Special populations' above and "Anticoagulation during pregnancy and postpartum: Agent selection and dosing" and "Venous thromboembolism in pregnancy and postpartum: Treatment" and "Anticoagulation therapy for venous thromboembolism (lower extremity venous thrombosis and pulmonary embolism) in adult patients with malignancy" and "Management of heparin-induced thrombocytopenia".)
●Duration – For patients with a first episode of acute VTE, the duration of anticoagulant therapy should be individualized according to the presence or absence of provoking events and risk factors, while taking into account the estimated risk for recurrence and bleeding and the individual patient's preferences and values (table 10). In general, the following applies:
•Minimum duration – For most patients with proximal DVT and/or symptomatic PE, we recommend anticoagulation for at least three months rather than for shorter periods (eg, four or six weeks) (Grade 1B). We also suggest three months of anticoagulation for distal DVT and isolated subsegmental PE (ISSPE) (Grade 2C).
Selected patients may additionally be candidates for longer finite treatment (eg, 6 to 12 months).
•Extending anticoagulation – Selected patients at high risk of recurrence (table 13) may benefit from longer periods of anticoagulation (ie, indefinite, no planned stop date). These data are discussed separately. (See "Selecting adult patients with lower extremity deep venous thrombosis and pulmonary embolism for indefinite anticoagulation".)
●Switching anticoagulants – Therapeutic anticoagulation should be ensured during transition periods. The strategy used depends upon the anticoagulants used for such periods. Interruptions during the first three months should be minimized due to the high risk of recurrent thrombosis during this period. (See 'Agents for treatment phase (long-term) anticoagulation' above and 'Switching anticoagulants during therapy' above.)
●Monitoring – All patients with acute DVT who are anticoagulated, especially those on factor Xa and direct thrombin inhibitors and those >75 years, should be monitored clinically for recurrence and bleeding, as well as for the signs and symptoms of conditions that may affect the half-life of the anticoagulant used (eg, kidney failure, weight gain or loss) and adverse effects of the medications (eg, skin necrosis, thrombocytopenia, osteoporosis). (See 'Monitoring' above.)
●Bleeding – Major bleeding rates on warfarin and LMW heparin during the first three months of anticoagulant therapy are low (less than 2 percent). Factor Xa and direct thrombin inhibitors have lower bleeding rates compared to warfarin; however, these rates may not reflect those in clinical practice. When reversal of anticoagulation is needed as part of a bleeding management strategy, the reversal strategy differs by agent. (See 'Risk of bleeding' above.)
ACKNOWLEDGMENT —
The UpToDate editorial staff acknowledges Russell D Hull, MBBS, MSc, who contributed to earlier versions of this topic review.
