Farmeconomia. Health economics and therapeutic pathways 2013; 14(3): 119-129

Original research

Venous thrombo-embolism (VTE) prevention of patients undergoing total hip and knee replacement: budget impact analysis of apixaban in Italy

Lorenzo G Mantovani 1, Giovanni Di Minno 1, Gianluca Furneri 2

1 Department of Clinical Medicine and Surgery, Federico II University of Naples, Italy

2 Fondazione Charta, Milan, Italy

Abstract

BACKGROUND: Venous thromboembolism (VTE) is a common and burdensome cardiovascular condition, frequently leading to severe complications and requiring high-cost healthcare interventions. New oral anticoagulants (nOACs) have demonstrated to be efficacious and safe in VTE prevention of patients undergoing total hip replacement (THR) and total knee replacement (TKR), a condition that is typically associated to cardiovascular disease. The Italian Healthcare Service (SSN) has recently approved the latest nOAC, apixaban. The present article aims to evaluate its economic impact in the perspective of the Italian SSN.

METHODS: We conducted a budget impact analysis to estimate clinical outcomes and economic consequences associated to the reimbursement of apixaban, in the prevention of VTE as a consequence of major orthopedic surgery, over a three-year time horizon. In our analysis we compared two alternative scenarios, with apixaban either reimbursed (Scenario B) or not reimbursed (Scenario A) by the Italian SSN, and estimated the difference of healthcare costs between the two scenarios. Only direct healthcare costs have been considered.

RESULTS: According to market assumptions, it is estimated that 1.2%, 3.7%, and 6.5% of THR patients, and 1.2%, 3.8% and 6.7% of TKR patients, would be treated with apixaban over the first three years since launch. At the estimated daily cost of apixaban (€2.48/die), this would translate into a budget impact of €14.3 mln, €45.5 mln, and €81.4 mln at years 1, 2 and 3 since launch, respectively. This expenditure would be more than offset by savings, due to: i) reduction of prescriptions of alternative treatment options (other nOACs, low-molecular weight heparins, fondaparinux); ii) reduction of the economic burden attributable of CV complications of VTE. Finally, Scenario B resulted slightly favourable compared to Scenario A, leading to economic savings for about €50 thousands over three years. Sensitivity analyses confirmed findings of the base-case analysis.

CONCLUSIONS: Reimbursement of apixaban does not determine a budget impact increase for Italian SSN. Its usage may be considered fully sustainable from a pharmaco-economic viewpoint.

Keywords

Apixaban, New oral anticoagulants, Venous thromboembolism

Corresponding author

Lorenzo G. Mantovani

Department of Clinical Medicine and Surgery Federico II University of Naples

Via Pansini 5, I-80131 Naples, Italy

mantovani@unina.it

Disclosure

Study funded by Bristol-Myers Squibb and Pfizer

Introduction

Venous thromboembolism (VTE) is the third most common cardiovascular illness after acute coronary syndrome and stroke [1], affecting more than 1 million patients each year [2]. It includes two different conditions, pulmonary embolism (PE) and deep vein thrombosis (DVT), which are relevant reasons of mortality and morbidity worldwide [3,4]. Without an efficacious prophylaxis, risk of developing VTE is estimated around 42-57% in patients undergoing total hip replacement (THP), and around 41-85% in patients undergoing total knee replacement (TKR) [5], with substantial implications in terms of healthcare costs and patients’ productivity for recurrent and prolonged inpatient stay [6-8].

In the last few years several studies have evaluated the economic impact of VTE. Recent estimates confirm that thromboembolism-attributable costs are more than $1.5 billion/year in the US [9]. In the same review, the authors highlighted that costs of acute venous episodes exceed $20,000 in certain patients’ categories. In a recent US study on Medicare beneficiaries [3], THR or THK patients developing VTE were found to have statistically significant higher costs, mortality and risk of rehospitalisation at one year after hospital discharge, compared to matched non-VTE patients. In a recent paper, Rupert et al. [10] reviewed ten economic studies carried out in the US and Europe, analyzing costs of THR and TKR patients developing VTE in the US and Europe. Evidence in this review pointed out that orthopaedic patients having PE and DVT costs up to more than twice than similar patients (controls) without any cardiovascular complications. Moreover, results from US administrative databases showed cost variability, with economic burden between US$3,000-9,500 for an initial episode. These costs seem higher than costs in the European Union, where VTE hospital management costs were €1,800 and €3,200 after 3 months and 1 year since diagnosis, respectively. Longer permanence in intensive coronary units, longer length stay, increasing usage of drug and outpatient services significantly increase direct healthcare costs of these patients [11]. Moreover, adopting a longer time horizon should take into account long-term clinical consequences of DVT such as recurrent DVT [12], post-thrombotic syndrome (PTS) [13] and thromboembolic pulmonary hypertension (PH) [14].

New oral anticoagulants (nOACs), apixaban, rivaroxaban and dabigatran, have demonstrated to reduce the risk of VTE after TKR and THR surgery, and to be as safe as low-molecular weight heparins (LMWH), enoxaparin in particular [15-22]. Although these new treatments have significantly improved patients’ outcomes in the last years, some concerns have been raised on their cost-effectiveness profile, due to the high daily costs. In particular, healthcare payers have expressed the need of evaluating the inclusion of nOACs in formularies in the light of the budget implications of their possible increasing usage. The present study aims to analyze the economic impact of reimbursing and funding apixaban for prevention of VTE after THR or TKR, from the perspective of the Italian healthcare system.

Material and methods

The present analysis evaluates the economic and clinical implications of reimbursing apixaban in Italy for VTE prevention in patients undergoing THR and TKR interventions, over the first three years after market authorization and reimbursement.

Model design and key analysis parameters

To perform this analysis, a decision-analytical budget impact model (BIM) running in Microsoft Excel® was developed. The model estimates financial impact (healthcare costs and investments) and clinical benefits (number of avoided VTE episodes) of alternative treatment scenarios which would likely reflect future trends of VTE prevention in Italy. Costs and outcomes for each hypothetical scenario have been evaluated on a short-time horizon (90 days). Figure 1 illustrates the decision tree built to run the analysis.

A patient undergoing THR or TKR is treated to prevent VTE events. During the follow-up period the patient could have a certain likelihood to develop different forms of fatal or non-fatal VTE (PE or DVT), as well as bleeding events with different severity. Distribution of patients within the flow, as well as overall management costs, depends on the drug received. In the budget impact analysis, two different scenarios were defined and compared to each other:

• “Apixaban not reimbursed” scenario (defined “Scenario A”), assuming that apixaban would not be reimbursed and funded by Italian National Healthcare Service (SSN);
• “Apixaban reimbursed” scenario (defined “Scenario B”), assuming that apixaban would be reimbursed and funded by Italian SSN.

Assumptions on patients distribution for both scenarios, in terms of drug prescription, were built using current data of patients’ treatment as reference, from updated market data [23]. According to current data, low molecular weight heparins represent the standard treatment for VTE prevention (administered to almost 90% of THR and TKR receiving VTE prevention, with enoxaparin being the most prescribed therapy). Fondaparinux and nOACs (rivaroxaban and dabigatran) are used in selected cases only. Table I shows the expected patients’ treatment distribution in the next three years after apixaban market authorization, according to scenario (apixaban reimbursed vs. apixaban not reimbursed) and indication (VTE prevention in THR and TKR).

For the present analysis, the perspective of the Italian SSN was adopted. The healthcare direct costs were measured over a 3-years period (with results split by year). Costs after the first year since apixaban market authorization in Italy were discounted at an annual rate of 3.5%, while clinical outcomes were not discounted. Costs were measured in Euro (€) 2012.

Patient population

In order to calculate the overall budget impact of VTE prevention in Italy, we estimated the annual number of patients receiving VTE pharmacological prevention due to THR and TKR interventions. Italian population at December 31st, 2010 (60.62 million inhabitants) and expected annual percentage increase in population (0.47%) were retrieved from the National Statistics Institute (ISTAT) [24]. A recent report, published in 2009 by the Superior Health Institute (ISS) [25], estimated that annual incidence of TKR and THR is 0.12% and 0.10% respectively. According to 2010 internal market research data [26], 96% of THR patients and 93% TKR patients would receive VTE pharmacological prevention. Assuming that percentage of treated patients would not change over time, the estimated annual number of treated patients at years 1, 2 and 3 (after apixaban market authorization) is 71.16, 75.08, 79.21 thousands for THR, and 63.37, 66.60 and 69.99 thousands for TKR.

Therapy duration and hospital length of stay data

Duration of therapy for apixaban and the other drugs indicated in VTE prevention has been estimated in order to calculate treatment costs. Table II shows treatment duration assumptions, by therapy, treatment setting, and type of replacement surgery.

In this budget impact model it was assumed that overall treatment duration for each drug was the same as in clinical trials submitted for EMA approval. According to clinical trials data, overall duration of treatment is similar for all drugs except dabigatran, which is shorter. It is assumed that treatment is initiated during inpatient stay (for the orthopaedic replacement intervention) and then continued at home after discharge. Distinction between hospital and domiciliary treatment was necessary, as cost of pharmacological treatments in Italy could be different in the two settings (cfr “Economic Data”). Assumptions on length of stay (by type of replacement surgery) were obtained from a study measuring resource consumption and economic burden of hip and knee replacement interventions in Italy [27].

Clinical data

Efficacy data for this analysis were derived from a mixed-treatment comparison evaluating evidence from clinical trials on VTE prevention in people undergoing elective THR and TKR surgery [28]. In absence of head-to-head clinical trials this metanalysis allowed to establish an indirect comparison of main clinical outcomes among VTE prevention treatments. This metanalysis was carried out according to NICE (National Institute of Clinical Excellence) requirements, and used as basis to build the economic evaluation of NICE single technology appraisal (STA) [29] Data for this metanalysis were selected through a systematic review, including the following clinical studies: double-blind randomised controlled trials (RCTs) of pharmacological prophylaxis of VTE in patients undergoing elective THR and TKR; trials enrolling mixed hip and knee populations, provided that patients undergoing hip surgery were randomised separately from patients undergoing knee surgery and patient numbers and results were stratified by surgery population; phase II trials, provided these were double-blind randomised controlled trials. Clinical evidence regarded LMWH, fondaparinux, rivaroxaban, dabigatran and apixaban, which could have been compared with either placebo or any of the treatments indicated in VTE prevention. Metanalysis measured two main clinical indicators: occurrence of VTE events (including pulmonary embolism, PE: symptomatic and asymptomatic deep vein thrombosis, DVT); occurrence of a bleeding events (major, non-major, minor events). In case of absence of calculated data for a certain drug, the rate of adverse events was assumed to be the lowest among compared treatments. Finally, rate of adverse events was stratified by distal vs. proximal DVT and fatal vs. non-fatal PE, using a conditional post-event distribution [28]. According to our assumptions, fatal events occurred in 12.5% and 25.0% of PE cases, for THR and TKR, respectively. Major bleedings represented 7.5% of all total bleeding events, for both THR and TKR patients. Table IIIa and Table IIIb illustrate, respectively, adverse events and death incidence rates (by drug and indication) and conditional distribution of post-event rates.

Economic data and assumptions on resource consumption

As the Italian SSN perspective was adopted in the analysis, only direct healthcare costs were measured. Overall VTE prevention therapy costs, inpatient costs for THR and TKR surgery, costs of VTE complications and bleedings were identified in our research. Figure 2 shows daily costs for pharmacological treatment.

With regard to apixaban, daily cost was calculated applying SSN rebates to the official ex-manufacturer price, published in the Official Gazette [30]. For the other therapies, a weighted average of inpatient and outpatient costs was used. Distinction between inpatient and outpatient costs was necessary as patients receiving initial treatment during hospitalization could continue therapy after discharge, if duration of treatment was longer than length of hospital stay (Table II). In our research, it was found that price of each marketed drug substantially varied among regions. Two main reasons determined this variability. First, Italian hospitals (or groups of hospitals) can negotiate prices with manufacturers through tenders, thus resulting in potentially different tender prices. Second, regions and local health units can discretionarily decide customized strategies to distribute heparins for outpatient use (i.e. direct distribution via hospitals, direct distribution via public pharmacies, standard retail distribution via public pharmacies), resulting again in different final prices for the SSN.

In order to obtain daily treatment costs being representative of the whole regional variability, we calculated average prices weighting regional prices by frequency of usage of a distribution channel and regional population. Regional prices and frequency of adoption of a certain distribution model were retrieved using multiple sources (internal market research data, official documentation on hospital tenders, etc.) Regional population data were finally extracted from Italian National Statistics sources (ISTAT data, December 31st 2010 [31]).

Costs for complications (VTE events and bleedings) requiring hospitalization are shown in Table III. In our model, only symptomatic VTE events determined healthcare costs for SSN. It was assumed that 100% of bleedings and 68.9% of symptomatic VTE events occurred at the time of hospital stay during the surgical replacement procedure [32]. In these cases an additional cost (40% of the DRG 78 tariff for PE events [33]; attributable DVT cost from PATHOS study [27]) was summed up to the standard hospital cost for THR or TKR (without complications) to take into account the higher burden for hospitals. A similar approach was used to estimate additional costs due to bleeding, by grade of severity. The remaining VTE events (31.1% of all cases) were assumed to occur after hospital discharge and to require hospital admission, according to costs and tariffs reported in Table IV.

Sensitivity analysis

A range of one-way deterministic sensitivity analyses was carried out to test robustness of base-case results. Specifically, variation of the following parameters was tested:

• Drug administration costs during THR/TKR hospitalization were included;
• Drug acquisition costs of nOACs were reduced of -40%, compared to base-case analysis;
• Discount rates for costs after year 1 were varied from 0% to 7% (discount in the base-case: 3.5%);
• Additional costs of VTE inpatient complications were varied of +10%, compared to base-case analysis;
• Percentage of patients receiving apixaban was varied of +10%, compared to apixaban patient share in the base-case.

Results

According to patient population and treatment assumptions, the estimated number of patients (THR and TKR) is 134,533 at Year 1, 141,672 at Year 2 and 149,190 at Year 3. In Scenario B, the estimated number of patients treated with apixaban would be 1,613 at Year 1, 5,309 at Year 2 and 9,839 at Year 3. Main results of the budget impact analysis are illustrated in Table V.

In Scenario A (i.e. apixaban not reimbursed by the SSN) the increasing number of THR and TKR interventions over time translates into a budget impact increase for the SSN (+1.81%: Year 2 vs. Year 1; +1.71%: Year 3 vs. Year 2); index hospitalizations represents the budget impact driver for the SSN, accounting for 98.83% of overall costs, considering the 3-year time horizon of the analysis. The budget impact associated to Scenario B (introduction of apixaban in the Italian market) and Scenario A (non-reimbursement scenario) would be similar: €3,641,168,854 and €3,641,223,204, respectively. Therefore Scenario B would determine cumulative cost savings for €54,350 vs. Scenario A in the first three years, with a cost-saving profile achieved in each of the three years analyzed. Also, the amount of cost savings sharply increases over time, as a consequence of the progressive replacement of LMWHs with nOACs.

Index hospitalizations represents the cost driver in the SSN perspective, being 98.83% of total direct healthcare costs. Excluding index hospitalization costs (for THR or TKR), drug costs account for 52.1% and 52.4% of total costs in Scenario B (apixaban reimbursed) and in Scenario A (apixaban not reimbursed), respectively (Figure 3). As a matter of fact, these costs (€22.24 million for Scenario A and €22.37 million for Scenario B) are relatively negligible, compared to inpatient costs for surgery replacement (€3.60 billion), which are the most relevant budget impact contributor of the analysis.

In Figure 4, the evolution of drug costs and VTE (inpatient and outpatient) costs in the first three years after apixaban reimbursement is illustrated. Drug investment with apixaban is fully set off by cost savings for avoided VTE episodes.

Based on clinical assumptions, Scenario B would be also associated to a reduction of VTE and bleeding events. In particular, over the entire period (3 years) apixaban introduction would be associated to the reduction of: symptomatic VTE and pulmonary embolism incidence (-106 non-fatal cases and -7 fatal cases vs. Scenario A); asymptomatic VTE incidence (-1,645 non-fatal cases vs. Scenario A); bleedings incidence (-186 cases vs. Scenario).

Results of sensitivity analysis, shown in Table VI, confirm that findings of the base-case study are robust.

In all scenarios tested, the inclusion of apixaban in the Italian market remains favourable. Inclusion of administration costs during index hospitalization, -40% reduction of drug costs, and +100% increase of apixaban usage would be the most beneficial scenarios, although the net benefit remains quite limited, if cost savings are compared with the overall budget impact of THR and TKR management, which is almost exclusively driven by hospital costs (+10% of hospital costs determines a +9.88% of overall budget impact for the SSN). The other scenarios included in the sensitivity analysis are affected by negligible fluctuations.

Discussion

In a few years, nOACs have become an important alternative to previous and well established pharmacological treatments indicated in VTE prevention after orthopaedic surgery. Currently LMWHs remain the standard of care/prevention in this indication, but the usage of nOACs has been rapidly increasing, due to the good risk-benefit profile, the mode of administration, and the daily costs, which are comparable with LMWHs and fondaparinux. This analysis highlights two important aspects regarding VTE prevention. First, the introduction of the new oral anti-coagulant apixaban would not determine a budget impact increase for the SSN. More precisely, the analysis shows that a net economic benefit, even if small, will be achieved in the next years with apixaban reimbursement and usage. Second, the economic impact of pharmacological prevention of VTE is relatively low, compared to the annual amount of hospital costs for THR and TKR interventions. VTE prevention is prescribed for about 15-30 days, according to type of prevention, type of drug and patients’ characteristics, and this sporadic usage makes the treatment affordable for the SSN. Our model estimates that only 1.7% of healthcare resources is allocated in VTE prevention, and that the introduction of new agents will not modify (increase) the number of treated patients, as almost all patients already receive a pharmacological VTE prevention.

In our analysis, we adopted a conservative approach for our base-case analysis. Drug administration costs (for parenteral treatments) were not included in the base-case analysis, assuming that these costs would be covered within the DRG (i.e. no extra-costs for hospitals). Also, we intentionally decided to focus on direct healthcare costs only, although the inclusion of indirect costs and the adoption of a societal perspective would have been favourable for apixaban and nOACs in general. Finally, and most importantly, apixaban daily cost only includes rebates granted to the official ex-manufacturer price, as the final result of government negotiations. Unlike the other treatments included in the analysis, further rebates granted to hospitals have not been estimated/included, due to the very recent decision of reimbursement. Reasonably, an additional reduction of apixaban costs would reflect into additional cost savings for the SSN.

The selection of alternative scenarios to test with sensitivity analysis was performed adopting different points of view. For assumptions affected by uncertainty (positive/negative), we evaluated standard sensitivity analysis. This was the case of discount rates and additional inpatient costs due to VTE complications. With regard of apixaban usage, the effect of a limit-case analysis, assuming to double apixaban patients’ share, was analyzed. Although this case may seem unlikely, it was tested to assess whether large, unexpected variations of the market could determine substantial budget impact increase, which would not be affordable, considering this period of economic constraints. Finally, the -40% reduction was chosen to take into account the future price erosion for nOACs at the time of launch of the second indication: prevention of stroke in patients with atrial fibrillation (SPAF). If the Italian Drug Agency (AIFA) will grant approval of nOACs in this new indication, a price reduction is likely to expect, and will be triggered by the following reasons: i) these drugs would be prescribed to a much larger number of patients, compared to the current situation; ii) pharmacological treatment SPAF is chronic; iii) likely, nOACs will be also distributed by public pharmacies, with additional costs for the SSN. All these reasons would lead to a substantial increase of economic resources to fund this class of agents, which will likely have to be offset by a price reduction of nOACs daily cost. Since the level of price discount for apixaban, rivaroxaban and dabigatran is hard to predict, we decided to apply a -40% reduction, which is the average final price rebate for enoxaparin (from ex-manufacturer price to the distribution price). This analysis shows that further discounts applied to nOACs would allow an additional slightly reduction of the budget impact, although majority of costs always depend on index hospitalizations.

The expected price reduction of nOACs in the next few years represents a definitive confirmation of affordability of nOACs usage in VTE prevention. On the other side, the possible approval and market authorization of these promising therapies in SPAF raise relevant questions on their cost-effectiveness, requiring further economic studies aimed to evaluate sustainability and comparative cost-effectiveness vs. available treatment options.

References

1. Goldhaber SZ. Pulmonary embolism thrombolysis: A clarion call for international collaboration. J Am Coll Cardiol 1992, 19: 246-7; http://dx.doi.org/10.1016/0735-1097(92)90473-Z

2. Dobesh PP. Economic burden of venous thromboembolism in hospitalized patients. Pharmacotherapy 2009; 29: 943-53; http://dx.doi.org/10.1592/phco.29.8.943

3. Baser O, Supina D, Sengupta N, et al. Clinical and cost outcomes of venous thromboembolism in Medicare patients undergoing total hip replacement or total knee replacement surgery. Curr Med Res Opin 2011; 27: 423-9; http://dx.doi.org/10.1185/03007995.2010.545940

4. Mantilla CB, Horlocker TT, Schroeder DR, et al. Frequency of myocardial infarction, pulmonary embolism, deep venous thrombosis, and death following primary hip or knee arthroplasty. Anesthesiology 2002; 96: 1140-6; http://dx.doi.org/10.1097/00000542-200205000-00017

5. Geerts WH, Bergqvist D, Pineo GF, et al.; American College of Chest Physicians. Prevention of venous thromboembolism: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition).Chest 2008; 133(6 Suppl): 381S-453S; http://dx.doi.org/10.1378/chest.08-0656

6. Vekeman F, LaMori JC, Laliberté F, et al. Risks and cost burden of venous thromboembolism and bleeding for patients undergoing total hip or knee replacement in a managed-care population. J Med Econ 2011; 14: 324-34; http://dx.doi.org/10.3111/13696998.2011.578698

7. Lefebvre P, Laliberté F, Nutescu EA, et al. All-cause and potentially disease-related health care costs associated with venous thromboembolism in commercial, Medicare, and Medicaid beneficiaries. J Manag Care Pharm 2012; 18: 363-74

8. Baser O, Supina D, Sengupta N, et al. Clinical and cost outcomes of venous thromboembolism in Medicare patients undergoing total hip replacement or total knee replacement surgery. Curr Med Res Opin 2011; 27: 423-9; http://dx.doi.org/10.1185/03007995.2010.545940

9. Dobesh PP. Economic burden of venous thromboembolism in hospitalized patients. Pharmacotherapy 2009; 29: 943-53; http://dx.doi.org/10.1592/phco.29.8.943

10. Ruppert A, Steinle T, Lees M. Economic burden of venous thromboembolism: a systematic review. J Med Econ 2011; 14: 65-74; http://dx.doi.org/10.3111/13696998.2010.546465

11. Chan CM, Shorr AF. Economic and outcomes aspects of venous thromboembolic disease. Crit Care Clin 2012; 28: 113-23; http://dx.doi.org/10.1016/j.ccc.2011.09.005

12. Kearon C. Natural history of venous thromboembolism. Circulation 2003; 107(23 Suppl 1): I22-30

13. Prandoni P. Healthcare burden associated with the post-thrombotic syndrome and potential impact of the new oral anticoagulants. Eur J Haematol 2012; 88: 185-94; http://dx.doi.org/10.1111/j.1600-0609.2011.01733.x

14. Piazza G, Goldhaber SZ. Chronic thromboembolic pulmonary hypertension. N Engl J Med 2011; 364: 351-60; http://dx.doi.org/10.1056/NEJMra0910203

15. Lassen MR, Raskob GE, Gallus A, et al.; ADVANCE-2 investigators. Apixaban versus enoxaparin for thromboprophylaxis after knee replacement (ADVANCE-2): a randomised double-blind trial. Lancet 2010; 375: 807-15; http://dx.doi.org/10.1016/S0140-6736(09)62125-5

16. Lassen MR, Raskob GE, Gallus A, et al. Apixaban or enoxaparin for thromboprophylaxis after knee replacement. N Engl J Med 2009; 361: 594-604; http://dx.doi.org/10.1056/NEJMoa0810773

17. Lassen MR, Gallus A, Raskob GE, et al.; ADVANCE-3 Investigators. Apixaban versus enoxaparin for thromboprophylaxis after hip replacement. N Engl J Med 2010; 363: 2487-98; http://dx.doi.org/10.1056/NEJMoa1006885

18. Lassen MR, Ageno W, Borris LC, et al.; RECORD3 Investigators. Rivaroxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty. N Engl J Med 2008; 358: 2776-86; http://dx.doi.org/10.1056/NEJMoa076016

19. Turpie AG, Lassen MR, Davidson BL, et al.; RECORD4 Investigators. Rivaroxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty (RECORD4): a randomised trial. Lancet 2009; 373: 1673-80; http://dx.doi.org/10.1016/S0140-6736(09)60734-0

20. Eriksson BI, Borris LC, Friedman RJ, et al.; RECORD1 Study Group. Rivaroxaban versus enoxaparin for thromboprophylaxis after hip arthroplasty. N Engl J Med 2008; 358: 2765-75; http://dx.doi.org/10.1056/NEJMoa0800374

21. Eriksson BI, Dahl OE, Rosencher N, et al.; RE-MODEL Study Group. Oral dabigatran etexilate vs. subcutaneous enoxaparin for the prevention of venous thromboembolism after total knee replacement: the RE-MODEL randomized trial. J Thromb Haemost 2007; 5: 2178-85; http://dx.doi.org/10.1111/j.1538-7836.2007.02748.x

22. Eriksson BI, Dahl OE, Rosencher N, et al.; RE-NOVATE Study Group. Dabigatran etexilate versus enoxaparin for prevention of venous thromboembolism after total hip replacement: a randomised, double-blind, non-inferiority trial. Lancet 2007; 370: 949-56; http://dx.doi.org/10.1016/S0140-6736(07)61445-7

23. IMS. Thromboembolism prevention in patients undergoing total hip or total knee replacement surgery in Italy: market data 2010. February 2011.

24. National Institute of Statistics. National Demographic Report, Year 2010. Available at: http://www.istat.it/it/archivio/28491

25. Healthcare Superior Institute (ISS). ISTISAN report 09/22. Published in November 2009

26. Bristol-Myers Squibb Internal Market Research Data: VTEp Knowledge book Italy, 2010

27. Degli Esposti L, Didoni G, Simon T, et al. Analysis of the cost of illness for the prevention of deep venous thrombosis in patients with total knee or hip replacement: results from the PATHOS study (Practice Analysis of THromboprophylaxis after Orthopaedic Surgery). Clinicoecon Outcomes Res 2013; 5: 1-7; http://dx.doi.org/10.2147/CEOR.S39978

28. Abacus international. Meta-analysis of apixaban and comparators for preventing venous thromboembolism in people undergoing elective knee and hip replacement surgery: NICE MTC analysis. November, 2010

29. National Institute of Clinical Excellence (NICE). Apixaban for the prevention of venous thromboembolism after total hip or knee replacement in adults: technology appraisal guidance 245. Issued: January 2012. Available at: http://www.nice.org.uk/nicemedia/live/13648/57895/57895.pdf

30. Italian Drug Agency (AIFA). Reimbursement and price of Eliquis (apixaban), authorized through European centralized procedure by European Commission. Official Gazette of the Italian Republic: August, 2012

31. National Institute of Statistics (ISTAT). Population in Italy, Year 2010. Available at: http://demo.istat.it

32. United Biosource. Resource consumption during total knee and hip replacement hospitalization. Last update: October 2012

33. Conference of Italian regions and and autonomous provinces. Unified conventional tariffs for hospital healthcare assistance (TUC), year 2009. Available at: http://www.regioni.it/it/show-doc_approvato_-_assistenza_ospedaliera_tariffa_unica_convenzionale_tuc_2009/news.php?id=104417

34. National Agency for Regional Healthcare Services (Agenzia Nazionale per i Servizi Sanitari Regionali; Age.Na.S). National and regional ambulatory/outpatient tariffs in 2009. Available at: http://www.agenas.it/monitoraggio_costi_tariffe/monitoraggio_costi_tariffe_prestazioni.htm

35. Nuijten MJ, Berto P, Kosa J, et al. Cost-effectiveness of enoxaparin as thromboprophylaxis in acutelly ill medical patients from the Italian NHS perspective. Recenti Prog Med 2002; 93: 80-91