Article
Oral anticoagulation for adults with atrial fibrillation or venous thromboembolism
- Aust Prescr 2025;48:161-6
- 14 October 2025
- DOI: 10.18773/austprescr.2025.047
Most patients with non-valvular atrial fibrillation (AF) or acute venous thromboembolism (VTE) can be treated with a direct-acting oral anticoagulant (DOAC); currently available DOACs are apixaban, rivaroxaban and dabigatran.
Warfarin is the first-line oral anticoagulant for valvular AF in patients with mechanical heart valves or rheumatic mitral stenosis.
Apixaban and rivaroxaban are first-line oral anticoagulants for cancer-associated VTE, and for AF or VTE in patients with body mass index over 35 kg/m2 or actual body weight over 120 kg.
All DOACs require dose adjustment in people with moderate kidney impairment.
Routine laboratory measurement of drug concentrations or relevant coagulation function assays is not required for safe and effective use of DOACs; however, there are situations when it may be beneficial, including emergency scenarios requiring normal haemostasis and where excessive or inadequate anticoagulation is suspected.
The oral anticoagulants approved for use in Australia are warfarin (a vitamin K antagonist) and 3 direct-acting oral anticoagulants (DOACs): apixaban and rivaroxaban (factor Xa inhibitors), and dabigatran (a direct thrombin inhibitor). Anticoagulants are effective at reducing harm and mortality from thromboembolic events (e.g. stroke), but are a major cause of iatrogenic harm from bleeding.1 This article focuses on the use of oral anticoagulants for atrial fibrillation (AF) and treatment of acute venous thromboembolism (VTE) in adults.
The choice of oral anticoagulant is driven by various patient characteristics, some of which are underpinned by the medicines' pharmacology (Table 1),2,3 and the presence of coexisting conditions such as chronic kidney disease, obesity and cancer.
Table 1 Characteristics of oral anticoagulants2,3
Warfarin | Direct-acting oral anticoagulants (DOACs) | |||
Apixaban | Rivaroxaban | Dabigatran [NB1] | ||
Mechanism of action |
Competitive inhibitor of vitamin K epoxide reductase |
Reversibly binding to the active site of factor Xa |
Reversibly binding to the active site of factor Xa |
Reversibly binding to the catalytic site of free and clot-bound factor IIa (thrombin) |
Dosing frequency for AF [NB2] |
Once daily |
Twice daily |
Once daily |
Twice daily |
Dosing frequency for VTE [NB2] |
Once daily |
Twice daily (higher dose for first 7 days) |
Twice daily for 21 days then once daily |
Twice daily (following 5 days of parenteral anticoagulant) |
Elimination half-life in healthy adults [NB3] |
40 hours |
10 hours |
7 hours |
14 hours |
Drug metabolism and transport pathways (examples of medicines inducing or inhibiting them) |
CYP2C9, CYP3A enzymes (rifampicin, carbamazepine, ritonavir, fluconazole) |
CYP3A enzymes, P-glycoprotein transporter (rifampicin, carbamazepine, ritonavir, ciclosporin, verapamil) |
CYP3A enzymes, P-glycoprotein transporter (rifampicin, carbamazepine, ritonavir, ciclosporin, verapamil) |
P-glycoprotein transporter (rifampicin, carbamazepine, ritonavir, ciclosporin, verapamil) [NB1] |
Key advantages compared with other oral anticoagulants |
Readily available anticoagulation intensity biomarker (INR) for dose adjustment Established safety in breastfeeding |
May have lowest bleeding risk in severe chronic kidney disease Standard doses sufficient for BMI over 35 kg/m2 or actual body weight over 120 kg |
Standard doses sufficient for BMI over 35 kg/m2 or actual body weight over 120 kg Once-daily dosing |
Readily available anticoagulation intensity biomarker (TCT) for detecting presence of anticoagulant |
Key disadvantages compared with other oral anticoagulants |
Routine periodic INR monitoring required Higher risk of intracranial haemorrhage Subject to more drug interactions |
Twice-daily dosing may reduce adherence Anticoagulant concentrations exceed safety threshold in breastmilk |
Doses of 15 mg and above need to be taken with food for optimal oral bioavailability |
Twice-daily dosing may reduce adherence Higher risk of anticoagulant accumulation with acute kidney injury compared with other DOACs Unclear if standard doses are sufficient for BMI over 35 kg/m2 or actual body weight over 120 kg |
Antidote |
Phytomenadione (vitamin K) |
Andexanet alfa [NB4] |
Andexanet alfa [NB4] |
Idarucizumab [NB4] |
AF = atrial fibrillation; BMI = body mass index; CYP = cytochrome P450; INR = international normalised ratio; TCT = thrombin clotting time; VTE = venous thromboembolism NB1: Dabigatran is formulated as the prodrug dabigatran etexilate, which is metabolised to the anticoagulant dabigatran after absorption. The prodrug, but not dabigatran, is subject to P-glycoprotein transport. NB2: Specifics regarding dosing amounts, including dose adjustment for renal impairment and drug interactions, can be found in resources such as the approved product information and the Australian Medicines Handbook. NB3: Half-lives of DOACs are significantly longer in the setting of renal impairment. NB4: The antidotes for DOACs have a limited role for reversal of anticoagulation in emergency situations.3 Refer to 'Management of bleeding in patients taking a DOAC', below, for details of their indications and risks. |
Randomised controlled trials and real-world data support the notion that warfarin and DOACs have similar efficacy for most patients requiring anticoagulation for AF or acute VTE;4-7 however, warfarin has a twofold higher risk of the uncommon complication of intracranial haemorrhage.5,8,9 Another disadvantage of warfarin is the high between- and within-patient variability in coagulation intensity, necessitating close monitoring with frequent laboratory coagulation testing via the international normalised ratio (INR).
At the time of writing, there are no published head-to-head randomised controlled trials directly comparing different DOACs, but publication of peer-reviewed findings from the recently completed COBRRA trial, which evaluated apixaban versus rivaroxaban for acute VTE, is awaited.7
Choice of anticoagulant for AF is unaffected in the setting of many cardiac valvulopathies, including the most common, aortic stenosis,10 and in the presence of bioprosthetic heart valves.11 However, there are 2 exceptions where warfarin appears superior to the other oral anticoagulants: mechanical heart valves and rheumatic mitral stenosis.12,13 In both of these conditions, randomised controlled trials found that vitamin K antagonists (mainly warfarin) were associated with a lower rate of ischaemic stroke than other anticoagulants.
Subcutaneously administered low-molecular-weight heparin, which inhibits factor Xa, has long been the mainstay of anticoagulation for cancer-associated VTE, with lower rates of recurrent thrombosis compared with warfarin. Oral factor Xa inhibitors (apixaban and rivaroxaban) are reasonable alternatives with similar efficacy and safety to low-molecular-weight heparin.14,15 There is insufficient data to support routine use of dabigatran in this setting.
Low-molecular-weight heparins are preferred over the oral anticoagulants for pregnant individuals, because of issues with either teratogenicity (warfarin) or a lack of safety data (DOACs).16
Warfarin remains the preferred anticoagulant in breastfeeding individuals, as it has the most safety data, with minimal anticoagulant in breastmilk. Dabigatran and rivaroxaban are emerging as reasonable oral anticoagulant alternatives, but apixaban is not recommended as breastmilk concentrations may exceed safety thresholds.17
All DOACs have some dependence upon the kidneys for clearance and require lower doses in the presence of kidney impairment. They are contraindicated below certain creatinine clearance values. Dabigatran has the greatest dependence upon kidney function and is contraindicated in patients with creatinine clearance below 30 mL/min. Rivaroxaban is contraindicated in patients with creatinine clearance values below 15 mL/min. Apixaban (dose-adjusted) is associated with a lower risk of major bleeding in severe chronic kidney disease than either warfarin or rivaroxaban, and may be preferred in patients with severe kidney disease.18
Warfarin has the least dependency upon the kidneys for clearance and, given the validity of INR as a biomarker to guide its dosing, some clinicians prefer warfarin in patients with severe kidney disease.
Liver disease is a consideration when prescribing anticoagulants due to the liver's central role in synthesising clotting factors and metabolising drugs.
Mild hepatic impairment (Child–Pugh A) does not significantly affect anticoagulant concentrations,2 and in the absence of existing coagulopathy, any anticoagulant may be used. Dosing should largely be informed by other factors such as kidney function.
In moderate to severe hepatic impairment (Child–Pugh B or C), anticoagulants are relatively contraindicated due to the risk of bleeding.19 Further, concentrations of apixaban, rivaroxaban and warfarin will be higher because these anticoagulants are subject to hepatic elimination. These patients should have their anticoagulation overseen by specialty services.
Efficacy and toxicity data from randomised controlled trials and real-world studies of therapeutic anticoagulation in patients with body mass index (BMI) over 35 kg/m2 or actual body weight over 120 kg support the use of rivaroxaban and apixaban at standard approved doses.20 These findings are consistent with studies of measured concentrations of rivaroxaban and apixaban in this patient group.21,22 There are insufficient data for dabigatran to provide reassurance for its efficacy at standard doses in patients with BMI over 35 kg/m2 or actual body weight over 120 kg; hence dabigatran is only advisable when supported by measurement of serum drug concentrations.23
For patients with low body weight (e.g. under 50 kg), there are insufficient data to support the use of standard DOAC doses, so DOAC use should be guided by a clinician with expertise in anticoagulation.24
INR-guided warfarin is reasonable to use in extremes of weight.20,24
All oral anticoagulants are susceptible to interactions with other medicines, although warfarin appears to be subject to more interactions than the others.25 Drug–drug interactions may lead to over-anticoagulation and risk of bleeding, or under-anticoagulation and risk of thrombosis.
A pharmacodynamic interaction that is common to all anticoagulants and may increase bleeding risk is concomitant use with an antiplatelet medicine.26 Such combinations are more justifiable in some situations (e.g. clopidogrel plus anticoagulant following percutaneous coronary intervention on a background of AF) than others (e.g. aspirin plus anticoagulant for primary cardiovascular disease prevention in a person with AF).27
Pharmacokinetic interactions are specific to the metabolic and transport pathways that are involved with each anticoagulant (Table 1).25,28 Several of the oral anticoagulants share common pathways, including cytochrome P450 3A enzymes that metabolise many medicines, and P-glycoprotein, an efflux transporter located on the luminal side of the gut wall and renal tubules.
Avoiding interactions that reduce anticoagulant efficacy is particularly important when treating an acute thrombotic event, as the risk from inadequate anticoagulation is high. For example, in a patient with a new deep vein thrombus already on long-term rifampicin (an inducer of several important drug-metabolising enzymes and transporters), a low-molecular-weight heparin (e.g. enoxaparin) may be preferred over an oral anticoagulant, at least initially.
In complex situations where a pharmacokinetic interaction cannot be avoided, a clinician with expertise in anticoagulation can be consulted about options such as anticoagulant dose adjustment based on the expected magnitude of the interaction and monitoring of anticoagulant concentrations (where available).28
In patients already on warfarin, switching to another oral anticoagulant is generally reasonable if the patient wants to switch and there are no contraindications.29,30 However, in frail older patients on warfarin with good INR control (time in therapeutic range of at least 70%), there are some data showing an increased risk of bleeding following a switch from warfarin.31
In patients on an oral anticoagulant who experience either a thromboembolic or bleeding event, a decision to switch may not always be rational. Before deciding to either switch anticoagulant or adjust the dose, it is important to diagnose the reason for the adverse event. For example, non-adherence contributes to recurrent thrombosis,32 and interactions with other medicines or with disease can alter risk of thrombosis or bleeding.25,28 Laboratory assays of drug concentration or coagulation function can be used to assist with diagnosing reasons for treatment failure.
While warfarin dosing is informed by regular INR testing, routine use of laboratory tests is not required for safe and effective use of DOACs in most patients. Nevertheless, there are 3 key clinical indications for laboratory DOAC measurement (Table 2).33-38
DOACs can be measured at the laboratory using either coagulation function assays (e.g. thrombin clotting time for dabigatran) or drug concentration assays (e.g. drug-specific anti-Xa assays for rivaroxaban and apixaban).33 When there is an indication for laboratory measurement, drug concentration assays are the preferred method, if locally available. When DOAC concentration assays are not available (either with timely turnaround or at all), coagulation screening assays may be a useful surrogate (Table 2).
Table 2 Indications for laboratory measurement of direct-acting oral anticoagulants (DOACs)2,33,34,36-38
Clinical indication | Examples of clinical scenarios | Comment about DOAC measurement |
Emergency scenarios requiring normal haemostasis |
|
Drug concentration assays: For all DOACs, drug concentration assays provide a precise gauge. A concentration below 30 micrograms/L may be used for proceeding with invasive procedures. Coagulation function assays: For apixaban and rivaroxaban, coagulation function assays are not recommended because of insufficient sensitivity to anticoagulant effect. For dabigatran, the coagulation function assay TCT has excellent sensitivity, and a result within normal reference range is consistent with absence of clinically significant dabigatran concentrations. APTT and PT are not useful. |
Scenarios suggestive of excessive DOAC exposure |
|
Drug concentration assays: Measure drug concentration and consult local laboratory or an anticoagulation expert for interpretation of results. Coagulation function assays: Coagulation function assays are difficult to interpret to provide a gauge of DOAC exposure. |
Scenarios suggestive of inadequate DOAC exposure |
|
As per Scenarios suggestive of excessive DOAC exposure |
APTT = activated partial thromboplastin time; CYP = cytochrome P450; PT = prothrombin time; TCT = thrombin clotting time |
The utility of drug concentration assays for assessing patient adherence to DOACs is limited, due to their short half-lives. For example, in a patient prescribed rivaroxaban once daily, taking a sample on Tuesday morning following a single missed dose on Monday (last dose 48 hours ago) will likely lead to an immeasurably low concentration that is indistinguishable from the patient who missed a week of doses.
When laboratory measurements are used to identify excessive or inadequate DOAC concentrations, interpretation of concentrations will vary depending on the timing of the sample relative to the last dose, and the anticoagulation indication. The local laboratory or a clinician with expertise in anticoagulation should be consulted for interpretation.
Assessing time since the last dose is crucial when managing a bleeding event in a patient using a DOAC, as the benefits of anticoagulation reversal may be negligible if most of the drug has already been eliminated. Plasma creatinine should be checked to gauge kidney function to inform estimates of DOAC clearance and half-life. In patients where the DOAC half-life is short, cessation of the anticoagulant is a sufficient intervention to reduce DOAC concentrations without resorting to antidotes.
Laboratory assessment of anticoagulation with dabigatran can be performed using the thrombin clotting time, as a normal value indicates negligible dabigatran effect.37 For apixaban and rivaroxaban, drug-specific anti-Xa assays can be used if available, as the coagulation screen assays have inadequate sensitivity for making a similar determination.37
Where ongoing anticoagulant effect is thought to be contributing to acute bleeding, a DOAC antidote can be considered, in consultation with a haematologist or clinical toxicologist, and depending upon local antidote availability and protocols. The antidotes idarucizumab and andexanet alfa function by specifically binding dabigatran and the factor Xa inhibitors (including enoxaparin), respectively. A limitation of andexanet alfa is its association with increased risk of thrombotic events, so it should be reserved for more severe bleeding events.3 A limitation of idarucizumab is its short half-life relative to dabigatran, and hence additional doses may be needed in some circumstances.39 Haemostatic agents such as prothrombin complex concentrate have a role for managing anticoagulant-associated bleeding where antidotes are not available.
Poor adherence to oral anticoagulants is associated with increased risk of thromboembolic events.32 This association appears more pronounced with DOACs than warfarin, plausibly because DOACs have shorter elimination half-lives and are less 'forgiving' of missed doses.
Once-daily dosing regimens may lead to greater adherence and persistence than twice-daily dosing.40 Other characteristics of the anticoagulants may also affect persistence. For example, dabigatran has more significant upper gastrointestinal adverse effects than other anticoagulants.40
Dose administration aids are often used to support adherence in patients taking multiple medicines. The manufacturer of dabigatran capsules warns against repackaging because this medicine has a tendency to absorb moisture, which can affect its stability. However, there is evidence that dabigatran can be repackaged in sealed blister packs and stored for 4 weeks at 25 C and 60% relative humidity.41-43
Oral anticoagulants are the mainstay of anticoagulant therapy for AF and VTE. While warfarin is the first-line anticoagulant for some specific scenarios, its use has declined because of its need for frequent monitoring of anticoagulation intensity. The DOACs, apixaban, rivaroxaban and dabigatran, have similar efficacy to warfarin and do not require routine monitoring. Knowledge of the individual characteristics of each oral anticoagulant can guide their safe and effective use.
This article was finalised on 16 September 2025.
Conflicts of interest: none declared
This article is peer reviewed.
Australian Prescriber welcomes Feedback.
Clinical Pharmacologist, Department of Clinical Pharmacology, Christchurch Hospital, Health New Zealand, New Zealand
Senior Lecturer, Department of Medicine, University of Otago, Christchurch, New Zealand
Clinical Pharmacologist, Department of Clinical Pharmacology, Christchurch Hospital, Health New Zealand, New Zealand
Associate Professor, Department of Medicine, University of Otago, Christchurch, New Zealand