Abstract
Cost-Effectiveness of Primary Prevention
There is now an extensive literature concerning cost-effectiveness of approaches commonly used for primary prevention of venous thromboembolism (VTE).1–40
In selecting and evaluating studies for this section, we include only those in which data for comparative effectiveness of approaches are based on randomized trials and/or systematic reviews of such trials, and which follow established guidelines for valid cost-effectiveness analysis.41–43 In this section, the perspective of analysis is that of the government health system or private insurance payer unless stated otherwise. In general, an approach is considered to be cost-effective if it is associated with an incremental cost per quality-adjusted life-year (QALY) of less than $ 50 000 or £20 000 to 30 000, which are thresholds commonly used to determine the society’s willingness-to-pay for health care interventions.44,45
In medium- and high-risk patients, the evidence establishes unequivocally that primary prevention with antithrombotic drugs or intermittent pneumatic compression is cost-effective compared to “no prophylaxis.”1–6,18,27 Primary prevention is also cost-effective compared to case finding (screening) for deep venous thrombosis (DVT). 2 Case finding does not prevent development of DVT and therefore does not reduce morbidity from PTS and its associated costs. Case finding is indicated in selected patients with contraindications to anticoagulant prophylaxis (eg, major trauma, see below). Data are not available for low-risk patients concerning cost-effectiveness for currently used prophylactic methods.
There have been several studies evaluating cost-effectiveness of primary prophylaxis using different anticoagulant drugs in patients having hip or knee replacement surgery or surgery for fractured hip.12,13,19,32,33 Two studies based on the US health care system,12,19 and one study based on the Norwegian system, 13 found that prophylaxis using fondaparinux was marginally less expensive than prophylaxis using enoxaparin. The Norwegian study found that the conclusions were sensitive to the price difference between the drugs and the type of surgery. A study based on the UK National Health Service found dabigatran etexilate was cost saving compared to enoxaparin 40 (40 mg once daily) in patients having total hip or knee replacement. 33 A study based on the Irish health care system evaluated cost-effectiveness of prophylaxis using either dabigatran etexilate or rivaroxaban compared to enoxaparin. 32 Finally, a study from the perspective of the Canadian health system found rivaroxaban to be a cost-effective alternative to enoxaparin. 38 Thus, the available evidence from studies in 3 different health systems indicates that both dabigatran and rivaroxaban are cost-effective alternatives to enoxaparin.32,33,38 The available evidence is inconclusive regarding the relative cost-effectiveness of rivaroxaban and dabigatran. 32
The cost-effectiveness of an extended duration of prophylaxis (28-35 days) after hip replacement or surgery for fractured hip has been evaluated in multiple studies.20,24,30,35,37 Two Canadian studies evaluated extended prophylaxis with low-molecular-weight heparin (LMWH) compared to warfarin or no extended prophylaxis.24,30 Dranitsaris et al 30 reported the incremental cost of 35 days of prophylaxis with dalteparin was Cdn $ 31 200 to 40 100 per QALY, whereas Skedgel et al 24 found an incremental cost of Cdn $ 106 454 per QALY for extended LMWH prophylaxis. The difference in these analyses may be explained by the proportion of patients requiring home-nursing services. The study by Dranitsaris appeared to assume no use of home nursing services 30 and Skedgel et al found extended prophylaxis with LMWH met the cost-effective threshold of Cdn $ 50 000 per QALY when less than 10% of the patients require home nursing services. 24 Two studies, one from Sweden 20 and the other from Italy, 35 both using a 5-year time horizon, suggest that fondaparinux is a cost-effective alternative to enoxaparin for extended prophylaxis and may be cost saving at 5 years. The Canadian study that found rivaroxaban to be cost-effective relative to enoxaparin in patients with hip replacement included a duration of prophylaxis of 35 days. 38
A limitation of applying these cost-effectiveness analyses is that they do not incorporate differences in values and preferences that may exist between surgeons or patients to avoid bleeding relative to preventing thromboembolism. Thus, an approach that increases bleeding, such as fondaparinux, even if found to be cost-effective or even cost saving, may not be accepted by surgeons or patients whose preferences are weighted to avoiding bleeding complications.
In patients with major trauma, although a regimen of the LMWH enoxaparin is more effective than unfractionated heparin for preventing DVT, an increase in major bleeding cannot be confidently excluded based on the results of the randomized trial comparing these approaches. 46 Cost-effectiveness modeling in this clinical scenario indicates that although enoxaparin appears to be a cost-effective alternative when considering the outcome of DVT averted, it is not cost-effective for the outcome of life-years gained because of the potential increase in major bleeding. 21 In patients with major trauma considered to have a contraindication to anticoagulant prophylaxis, combined short-term (2 weeks) intermittent pneumatic compression and case finding with serial Doppler ultrasonography for the duration of hospitalization is more cost-effective than prophylactic placement of an inferior vena cava filter. 29
Using the perspective of US Medicare reimbursement, Heerey and Suri 14 evaluated the cost-effectiveness of 2 regimens of LMWH (dalteparin 5000 or 2500 U daily) compared to unfractionated heparin for primary prevention in patients undergoing abdominal surgery. The base-case analysis suggested that both dalteparin regimens were cost-effective using an incremental cost-effectiveness threshold of $ 50 000 per QALY gained. 14 However, sensitivity analysis indicated that there was substantial uncertainty in the cost-effectiveness results, in part due to the influence of patient age and gender. In the base analysis, unit costs for the dalteparin 2500 and 5000 U regimens were more than 10 and 20 times that of unfractionated heparin. 14 Sensitivity analysis showed that reducing the cost of dalteparin by 50% would result in the 2500 U regimen being the more cost-effective, and the 5000 U regimen would be cost-effective by comparison to either the 2500 U dalteparin or the unfractionated heparin. Thus, in health care systems in which the cost of LMWH is much lower relative to unfractionated heparin than in the United States, primary prevention using LMWH in patients having abdominal surgery may have acceptable incremental cost-effectiveness or may even be the most cost-effective, depending on the regimen.
The cost-effectiveness of primary prevention in hospitalized medical patients using LMWH or unfractionated heparin has been evaluated in 4 studies10,18,22,27; the health system was in the United States in 3 studies10,22,27 and in Germany in 1 study. 18 The results of all 4 studies are consistent indicating that prophylaxis with LMWH is more effective and less costly than with unfractionated heparin.
The cost-effectiveness of primary prevention of VTE during pregnancy using once daily LMWH in women with a single previous episode of VTE has been evaluated. 15 The results indicate that primary prevention is cost-effective for “high-risk” women with a prior idiopathic VTE or a known thrombophilic condition if the risk of bleeding is 1% or lower.
Cost-Effectiveness of Secondary Prevention (Treatment to Prevent Recurrent VTE)
The criteria for selecting studies to evaluate cost-effectiveness of alternative approaches for secondary prevention included randomized trials and/or systematic reviews of such trials to determine comparative effectiveness and established guidelines for cost-effectiveness.41–43 However, most studies to date have not used the QALY as the measure of effectiveness, and conclusions from these studies are based on cost per event of recurrent VTE.
The current standard care for most patients with established DVT or pulmonary embolism (PE) is anticoagulation consisting of initial treatment with either LMWH or intravenous unfractionated heparin followed by long-term treatment with a vitamin K antagonist (eg, warfarin). The cost-effectiveness of anticoagulant therapy has been formally evaluated.47–50 The cost-effectiveness of other approaches such as intravenous thrombolytic therapy, catheter-directed thrombolytic therapy and/or thrombus removal, or insertion of a vena cava filter has not been evaluated; these approaches have usually been reserved for specific indications in selected patients.
Two studies have compared the cost-effectiveness of intravenous unfractionated heparin with subcutaneous LMWH for the initial treatment of patients with DVT.47,48 The findings are consistent and indicate that LMWH is cost-effective. Hospitalization is the major driver for cost.47,48 The LMWH is an effective approach to treat DVT out of hospital.51,52 The LMWH for initial therapy is a cost-saving approach if 8% or more of patients are treated entirely as outpatients or 13% or more have a reduced hospital stay. 47
Long-term anticoagulation is required in patients with VTE to prevent recurrent thromboembolism. The standard approach has been treatment with a vitamin K antagonist with the dose adjusted according to laboratory monitoring of the anticoagulant effect. Long-term therapy with a vitamin K antagonist is highly effective and is cost-effective compared to inadequate long-term therapy. 50 However, the need for laboratory monitoring is associated with significant costs49,50 and is a burden that influences quality of life in many patients. Approaches to improve the effectiveness, safety, and efficiency of oral VKA therapy include specialized anticoagulation clinics and patient self-monitoring. The data on cost-effectiveness of these approaches in patients with VTE is limited, since the studies have included a mixed population with various indications for long-term therapy (eg, heart valves, atrial fibrillation, etc). The UK Health Technology Assessment Programme concluded that patient self-monitoring is unlikely to be more cost-effective than specialized anticoagulation clinics, using a threshold of £30 000 per QALY, 49 although patient self-monitoring may improve quality of life for some patients who travel frequently or have difficulty traveling to the clinic. 49
The LMWH therapy given in fixed doses without anticoagulant monitoring is an effective and safe approach to treat VTE for 3 to 6 months,53–55 but the cost-effectiveness of 3 to 6 months therapy with LMWH has not been formally evaluated. The LMWH is preferred in patients with cancer with VTE because it is markedly more effective than VKA treatment (number needed to treat to prevent one recurrent VTE of approximately 13).53,54 The LMWH is also effective in the broad spectrum of VTE patients without cancer, and in such patients, is associated with improvement in the patient’s perceived quality of life. 55
The new oral anticoagulants dabigatran and rivaroxaban have been evaluated for treatment of VTE including long-term therapy for 3 to 12 months.56,57 These drugs do not require laboratory monitoring of the anticoagulant effect and therefore, greatly simplify long-term anticoagulant treatment, but their cost-effectiveness remains to be evaluated.
The role of laboratory screening for thrombophilia in guiding clinical decisions about an extended or indefinite duration of anticoagulant therapy has garnered much debate. The UK Health Technology Assessment Programme concluded that scenarios were found where such an approach is cost-effective using a threshold of £ 20 000 per QALY, but the results are subject to significant uncertainty because of a lack of randomized trials or definitive data on the magnitude of increased risk of recurrence for different categories of thrombophilia. 58 The relative cost-effectiveness of routine screening for thrombophilia versus targeted screening based on patient and family history requires further studies. 58
A recent meta-analysis indicates that patients with unprovoked (idiopathic) proximal DVT or PE have a high annual risk of recurrence whenever treatment is stopped, whether the duration of treatment is 3, 5, 12, or 27 months. 59 This finding has important implications for future cost-effectiveness analysis.