Chronic coronary artery disease (CAD) and peripheral artery disease (PAD) are manifestations of atherosclerosis that result from the narrowing and blockage of the coronary and limb arteries, respectively.1,2 CAD is the most common form of cardiovascular (CV) disease and is a leading cause of death in the United States.3 Patients with CAD can have complications such as angina, and are at an increased risk for stroke and myocardial infarction (MI) compared with patients without CAD.1,4 The prevalence of PAD increases with age, and affects approximately 20% of Americans aged >55 years, and the incidence of PAD is higher in men than in women.5 Patients with PAD typically suffer from unusual leg pain, claudication, and/or critical limb ischemia.6
The current management of CAD and PAD includes lifestyle modifications (eg, dietary changes, smoking cessation, and exercise),7 pharmacologic treatments (eg, beta blockers and/or angiotensin-converting enzyme inhibitors for blood pressure control, statins and proprotein convertase subtilisin/kexin type 9 inhibitors to lower cholesterol, and antiplatelet therapy, including low-dose aspirin and clopidogrel),7-10 medical devices (eg, pacemakers, coronary stents), and procedures (eg, bypass surgery, angioplasty).8,9,11-13 Notwithstanding the treatments and interventions available, a significant unmet need exists for more effective treatment options in patients with CAD or PAD because of the high risk for CV events and death.14-18
CAD and PAD each pose a significant economic burden to the healthcare system and society. The American Heart Association estimated that CAD is associated with annual direct medical costs of $89 billion and indirect costs (from productivity loss) of $99 billion19; the annual direct medical costs for PAD were estimated to be more than $21 billion.20 A key contributor to such costs is the disease burden from major adverse CV events (MACE) and major adverse limb events (MALE).18,21 Although the exact contributions of MACE and MALE to the total costs of CAD and PAD is unclear, a recent analysis of healthcare claims and electronic medical records showed that patients with CAD and/or PAD who had MACE and/or MALE averaged an additional $44,495 in 1-year healthcare costs versus patients with CAD and/or PAD without these events.18
The efficacy and safety of rivaroxaban, a direct inhibitor of factor Xa, was assessed in patients with chronic CAD and/or PAD in the phase 3, multinational, double-blind, randomized, placebo-controlled Cardiovascular Outcomes for People Using Anticoagulation Strategies (COMPASS) clinical trial.22 The rate of MACE (a composite primary end point of CV death, MI, or stroke) was significantly lower in patients who received rivaroxaban 2.5 mg in combination with aspirin 100 mg than in patients who received monotherapy with aspirin 100 mg (hazard ratio [HR], 0.76; 95% confidence interval [CI], 0.66-0.86; P <.001).22
Similarly, a reduction in the rate of MALE (a composite of chronic and acute limb ischemia, and major amputation that resulted from vascular events) was observed in patients who received rivaroxaban plus aspirin compared with aspirin alone (HR, 0.54; 95% CI, 0.35-0.82; P = .0037).23,24 The COMPASS study was stopped as a result of overwhelming evidence of efficacy in 2017 (mean follow-up, 23 months), when only half of the planned primary end point events to detect a 20% lower risk for MACE in each of the 2 comparisons of rivaroxaban versus aspirin had occurred.22
In addition, rivaroxaban could substantially reduce the costs associated with MACE and MALE in patients with CAD and/or PAD. A recent cost-effectiveness analysis of rivaroxaban plus aspirin versus aspirin alone from the UK National Health System perspective demonstrated that rivaroxaban plus aspirin is a cost-effective alternative to aspirin alone over a lifetime horizon in patients with chronic CAD and/or PAD.25
To explore such possibilities, the aim of the current study was to estimate the 1-year economic implications of preventing MACE and MALE with the use of rivaroxaban plus aspirin versus aspirin alone for the treatment of patients with chronic CAD and/or PAD who are members of a commercial health plan in the United States.
We developed a cost-consequence model in Microsoft Excel 2016. The model compared a treatment scenario without rivaroxaban in which all patients with chronic CAD and/or PAD in a hypothetical commercial health plan of 1 million members received aspirin 100 mg alone, with a scenario in which, on the introduction of rivaroxaban, a proportion of patients in the same health plan would receive rivaroxaban plus aspirin (Figure).
For both scenarios, the model calculates the total number of clinical events, including MACE, MALE, and major bleeding events, that would occur over 1 year, as well as the direct medical costs, including drug acquisition costs and the healthcare costs associated with the incidence of clinical events. The differences in the total number of clinical events and the total costs between the 2 scenarios were used to estimate the net clinical and economic implications of using rivaroxaban plus aspirin versus aspirin alone.
A description of the methodology for the analysis used to inform this model was published previously.17 Briefly, analyses of claims and electronic medical records were conducted using data from January 1, 2009, to September 30, 2016, from the Optum Integrated Database to identify patients with a diagnosis of chronic CAD and/or PAD.18 Among 3,881,594 enrollees, the final study sample of 99,730 included 43,555 patients aged <65 years with chronic CAD and/or PAD who were considered for this analysis.18 The model inputs derived from the Optum Integrated Database analyses are shown in Table 1.
The number of patients with chronic CAD and/or PAD who were eligible for treatment in the model scenarios was calculated by scaling the number of patients with chronic CAD and/or PAD identified in the Optum Integrated Database to a hypothetical commercial health plan of 1 million members (N = 11,220 = 43,555/3,881,594 × 1 million).
Our analytical cohort was based on the definition of CAD and PAD used in the COMPASS study. Patients with nonvalvular atrial fibrillation were excluded in our study, because our focus was on the applicability of COMPASS to clinical practice, and rivaroxaban was indicated for nonvalvular atrial fibrillation at the time of this analysis.18 The matching of patient characteristics was not performed between the real-world evidence cohort and the population in COMPASS to identify the real-world burden of cost and the incidence of clinical events.
The plan size (ie, 1 million members) was assumed to remain constant over the 1-year time horizon.
The definitions used to identify patients with chronic CAD and/or PAD who had MACE and/or MALE in the Optum Integrated Database were consistent with those used in the COMPASS study.18,22 The algorithm validated by Cunningham and colleagues was used to identify patients who had major bleeding events.27
In our model, patients who received aspirin alone in both scenarios, with and without rivaroxaban, were assumed to have the same incidence rates of clinical events as patients from the Optum analytical cohort (mean follow-up, 1.8 years; Table 1). The incidence rates of clinical events for patients receiving rivaroxaban plus aspirin were calculated by applying the relative efficacy of rivaroxaban plus aspirin versus aspirin alone (ie, HR; Table 1) from the COMPASS study to the incidence rates of clinical events that were observed in the Optum analytical cohort.22
The reduction of MALE in patients who received rivaroxaban plus aspirin versus patients who received aspirin alone was derived from the subgroup analyses of the COMPASS population with PAD, which was assumed to be similar for the model target population (ie, patients with chronic CAD and/or PAD).22,23 To calculate the total number of clinical events in each scenario over the model time horizon, the number of patients with chronic CAD and/or PAD who received rivaroxaban plus aspirin or aspirin alone was multiplied by the treatment-specific incidence rates of each clinical event.
To calculate the healthcare costs associated with the clinical events of interest from the Optum Integrated Database, patients were stratified based on whether they had an event of interest.18 A total of 5 interactions of clinical events were created in the analytical cohort to inform the model, including MACE with or without major bleeding, MALE with or without major bleeding, and major bleeding without MACE or MALE.
Multivariate linear regression models were developed to compare the total 1-year healthcare costs (defined as the sum of the inpatient, outpatient, and prescription pharmacotherapy costs) of patients who had an event of interest versus patients who did not. The costs associated with rehabilitation after critical limb ischemia, amputation, and stroke were not fully captured, given the 1-year analysis time frame.
The covariates used in these models included age, sex, race, payer type, smoking status, body mass index, preindex coronary artery bypass graft or percutaneous coronary intervention, limb ischemia, dyslipidemia, hypertension, and Charlson Comorbidity Index.18
A difference in annual healthcare costs between the patients who had an event and those who did not have an event was assumed to be attributable to the event of interest, and therefore was used in the model (Table 1). All the costs were derived based on the reimbursement amount reported in the Optum Integrated Database and were adjusted to 2019, using the medical care component of the US Consumer Price Index.28
The model included the 2019 wholesale acquisition costs (WACs; obtained from the RED BOOK) of aspirin and rivaroxaban, which were $0.27 and $447.99, respectively, for a 30-day supply.26 The annual drug acquisition costs were calculated by multiplying the 30-day WACs by 12, resulting in $3.24 for aspirin and $5375.88 for rivaroxaban plus aspirin. In the base-case analysis, a 25% rebate was assumed for the WAC of rivaroxaban, which resulted in an annual cost of $4035.12 for rivaroxaban plus aspirin.
Our assumption for the 25% rebate is lower than the 50% rebate assumed in the incremental cost-effectiveness ratio assessment of additive CV treatments, including rivaroxaban. Hence, we believe that our assumption of rebate here represents a conservative approach. As a conservative approach, no treatment discontinuation was modeled, and all patients were assumed to be 100% adherent to their medications, which maximized the drug acquisition costs for both scenarios.
In the base-case analysis, 5% of the target population was assumed to have received rivaroxaban plus aspirin and the remaining 95% was assumed to have received aspirin alone in the scenario with rivaroxaban. The reported findings for each scenario include the number of clinical events (ie, MACE, MALE, and major bleeding), treatment acquisition costs, healthcare costs associated with the incidence of clinical events, total cost, and the total cost per member per month (PMPM), which was calculated for each scenario by dividing the total cost to the health plan by the 1 million members in the plan and the 12 months in the model time horizon. The incremental total cost and incremental PMPM cost were also reported.
One-way sensitivity analyses were conducted to assess the impact of varying key model parameters (20% rebate for rivaroxaban acquisition cost, rivaroxaban market share at 10% and 20%, HRs of rivaroxaban plus aspirin based on their reported 95% CIs in the COMPASS study, and incidence rates and healthcare costs of clinical events [±20% variation was assumed]) on the incremental PMPM cost. The parameter values used in the one-way sensitivity are shown in Supplementary Table S1.
In the subgroup analysis, the clinical and economic implications of using rivaroxaban plus aspirin versus aspirin alone in patients with chronic CAD and/or PAD who had ≥2 risk factors of MACE and/or MALE were also analyzed. Multivariate Cox proportional hazards models were developed using the Optum Integrated Database to identify the significant (P ≤.01) risk factors of MACE and/or MALE. These risk factors were congestive heart failure, diabetes, previous stroke, cerebrovascular disease, atrial fibrillation, peripheral vascular disease other than PAD, dementia, chronic pulmonary disease, renal disease or dysfunction, and being a smoker. The model inputs specific to this subgroup are shown in Table 1.
In the scenario without rivaroxaban, 11,220 patients with chronic CAD and/or PAD received treatment with aspirin alone for 1 year (Table 2). With the introduction of rivaroxaban, 561 of these patients received treatment with rivaroxaban plus aspirin (5% market share), and the remaining 10,659 patients received aspirin alone. The introduction of rivaroxaban resulted in a reduction of MACE and MALE compared with aspirin monotherapy.
These reductions translated to annual total cost-savings of $537,174 ($196,816 and $340,358 in MACE- and MALE-associated costs, respectively). Savings in MACE- and MALE-associated costs partially offset the additional costs of major bleeding events and drug acquisition ($108,737 and $2,294,809, respectively) with rivaroxaban plus aspirin, resulting in an increase in total costs of $1,866,372 ($0.16 PMPM) compared with aspirin alone (Table 2).
The base-case incremental PMPM cost was the most sensitive to change in the market share of rivaroxaban plus aspirin (Supplementary Table S1). The model results were robust to changes in other model parameters, such as the rebate for rivaroxaban’s drug acquisition cost, the rates of clinical events, the treatment effect of rivaroxaban plus aspirin, and the annual healthcare cost of clinical events.
The results of the subgroup analysis are presented in Table 3. The incremental PMPM cost for 1 year of treatment was $0.09 in patients with ≥2 risk factors of MACE and/or MALE. The decrease in PMPM cost compared with the base-case results was a consequence of an increase in the offset of rivaroxaban’s acquisition costs by savings generated from the greater reduction in the incidence of MACE and MALE in patients who received treatment with rivaroxaban plus aspirin.
At a 5% market share, treatment with rivaroxaban plus aspirin was associated with an incremental PMPM cost of $0.16 during 1 year for a US commercial health plan of 1 million members. Rivaroxaban’s acquisition cost was partially offset by savings from a reduction in the incidence of MACE and MALE with rivaroxaban plus aspirin versus aspirin alone. The market share of rivaroxaban plus aspirin was the most impactful parameter on the model’s results, with changes in the incremental cost PMPM being directly proportional to changes in the value of the market share. However, the impact of this parameter on the model results should be viewed in the context of low-cost generic aspirin as a comparator.
In the Optum analytical cohort, approximately 50% of the population had ≥2 risk factors of MACE and/or MALE. In this subgroup, the incremental PMPM cost of rivaroxaban plus aspirin was approximately 45% lower than the base case because of the increased offset of rivaroxaban’s acquisition cost by savings generated from the greater reductions in MACE and MALE in patients at higher risk for these clinical events versus patients with any (ie, ≥0) risk factors of MACE and/or MALE.
A post hoc analysis of patients with a high risk for vascular events from the COMPASS clinical trial demonstrated that rivaroxaban plus aspirin’s net clinical benefit remains favorable in this subgroup, which supports our findings.29 To our knowledge, no other economic evaluations in US patients with chronic CAD and/or PAD are published from the US payer perspective to compare the model results. (As noted earlier, a similar analysis was conducted in the United Kingdom.25)
Our analysis is conservative in 2 main aspects. The model assumptions of 100% adherence to treatment, no treatment discontinuation, and no patient cost-sharing result in conservative estimates of the incremental cost of rivaroxaban plus aspirin to a health plan. In addition, the prevention of MACE and MALE can lead to the possible avoidance of long-term events, such as limb amputations, especially in patients with PAD.
The 1-year time horizon was considered appropriate for US commercial payers for whom their insured population might change every year. A 1-year time horizon is common in a US budget impact analysis (which is similar to our analysis) of chronic conditions. This has been highlighted in a methodological review of US budget impact models by Mauskopf and Earnshaw.30
This study has several limitations. First, the model did not account for the impact of treatment on the quality of life of patients. In addition, we assumed that the prevalence of patients with chronic CAD and/or PAD will remain constant over the analyses’ time horizon, and we did not include non-CV mortality in our analyses.
Also, the rates of clinical events were assumed to be time-independent, and the model did not consider the impact of previous event history on the risk for future or subsequent clinical events. However, given the short time horizon of the analyses, we believe that the net impact of these assumptions is minimal and does not significantly affect the results of the analysis.
Furthermore, the safety and efficacy estimates observed in the controlled setting of the COMPASS trial may not accurately estimate the safety and efficacy of rivaroxaban plus aspirin in the real world. This limitation is applicable to all economic evaluations that use clinical trial–based estimates.
Because of the 1-year time horizon of our analyses, we were not able to capture the potential downstream cost offsets, which may lead to the underestimation of the economic value of rivaroxaban plus aspirin.
And because COMPASS was a multinational clinical trial, the efficacy and safety estimates are from multiple countries and are not only from the US population. However, as noted in the sensitivity analyses, our results were robust, despite variations in these efficacy and safety estimates.
Finally, major bleeding in our analysis was defined using the algorithm by Cunningham and colleagues,26 given its validity for use in electronic healthcare databases, such as the Optum Integrated Database. In the COMPASS trial, major bleeding was defined using the modified International Society on Thrombosis and Haemostasis criteria, which included fatal bleeding, symptomatic bleeding into a critical organ, bleeding into a surgical site requiring reoperation, and bleeding that led to hospitalization.22 However, the 2 sets of definitions are broadly consistent and are expected to have minimal impact on the model’s results.
Our study estimated the real-world economic and clinical impact of outcomes observed in the COMPASS clinical trial, which combined the strengths of a randomized controlled trial and an observational database rather than using data collected from randomized controlled trials under ideal circumstances only, to generate policy-relevant results and to inform formulary decision makers regarding the inclusion of rivaroxaban plus aspirin in the treatment of patients with chronic CAD and/or PAD. The findings from this study have implications for coverage decisions in many health plans in which antiplatelet therapy is regarded as a treatment option for patients with chronic CAD and/or PAD.
The results of this study suggest that antiplatelet therapy may lead to more costs associated with the management of MACE and MALE and, potentially, MACE- and MALE-related long-term events (eg, limb amputations) and their associated management and rehabilitation costs. In a new era of emerging thrombocardiology, rivaroxaban plus aspirin offers an effective thrombotic risk management strategy for healthcare providers, payers, and patients in the management of chronic CAD and/or PAD.
With the current and likely growing disease and economic burden of MACE and MALE among patients with chronic CAD and/or PAD, the contribution of rivaroxaban would be increasingly important if it can be widely accessible and by identifying high-risk patients for higher cost-savings potential.
Funding for this study was provided by Janssen Scientific Affairs.
Author Disclosure Statement
Mr Hernandez was an employee of Evidera, a consulting company, when the study was conducted. Mr Shah is an employee of Evidera. Dr Zhao is a former employee of, and Dr Milentijevic and Dr Kharat are employees of Janssen Scientific Affairs.
- Sanchis-Gomar F, Perez-Quilis C, Leischik R, Lucia A. Epidemiology of coronary heart disease and acute coronary syndrome. Ann Transl Med. 2016;4:256.
- National Institutes of Health. Peripheral artery disease: also known as P.A.D. 2019. www.nhlbi.nih.gov/health/health-topics/topics/pad. Accessed October 22, 2019.
- Benjamin EJ, Blaha MJ, Chiuve SE, et al; for the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart Disease and Stroke Statistics—2017 Update: a report from the American Heart Association. Circulation. 2017;135:e146-e603. Errata in: Circulation. 2017;135:e646; Circulation. 2017;136:e196.
- Centers for Disease Control and Prevention. Conditions that increase risk for stroke. www.cdc.gov/stroke/conditions.htm. Accessed September 11, 2020.
- Hankey GJ, Norman PE, Eikelboom JW. Medical treatment of peripheral arterial disease. JAMA. 2006;295:547-553.
- Hirsch AT, Haskal ZJ, Hertzer NR, et al. ACC/AHA 2005 guidelines for the management of patients with peripheral arterial disease (lower extremity, renal, mesenteric, and abdominal aortic): executive summary. J Am Coll Cardiol. 2006;47:1239-1312.
- Smith SC Jr, Benjamin EJ, Bonow RO, et al. AHA/ACCF secondary prevention and risk reduction therapy for patients with coronary and other atherosclerotic vascular disease: 2011 update: a guideline from the American Heart Association and American College of Cardiology Foundation. Circulation. 2011;124:2458-2473. Erratum in: Circulation. 2015;131:e408.
- Singh P, Harper Y, Oliphant CS, et al. Peripheral interventions and antiplatelet therapy: role in current practice. World J Cardiol. 2017;9:583-593.
- Aboyans V, Ricco JB, Bartelink MLEL, et al. 2017 ESC Guidelines on the Diagnosis and Treatment of Peripheral Arterial Diseases, in collaboration with the European Society for Vascular Surgery (ESVS): document covering atherosclerotic disease of extracranial carotid and vertebral, mesenteric, renal, upper and lower extremity arteries. Eur Heart J. 2018;39:763-816.
- Levine GN, Bates ER, Bittl JA, et al. 2016 ACC/AHA guideline focused update on duration of dual antiplatelet therapy in patients with coronary artery disease: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2016;134:e123-e155. Erratum in: Circulation. 2016;134:e192-e194.
- Wilkins LR, Sabri SS. Strategies to approaching lower limb occlusions. Tech Vasc Interv Radiol. 2016;19:136-144.
- Shishehbor MH, Jaff MR. Percutaneous therapies for peripheral artery disease. Circulation. 2016;134:2008-2027.
- Topfer LA, Spry C. New technologies for the treatment of peripheral artery disease. In: CADTH Issues in Emerging Health Technologies; issue 172. Ottawa, Canada: Canadian Agency for Drugs and Technologies in Health; April 2018. www.cadth.ca/sites/default/files/pdf/eh0064-peripheral-artery-disease.pdf. Accessed November 5, 2019.
- Centers for Disease Control and Prevention. Heart disease facts. www.cdc.gov/heartdisease/facts.htm. Accessed November 5, 2019.
- Marrett E, daCosta DiBonaventura M, Zhang Q. Burden of peripheral arterial disease in Europe and the United States: a patient survey. Health Qual Life Outcomes. 2013;11:175.
- Shanmugasundaram M, Ram VK, Luft UC, et al. Peripheral arterial disease—what do we need to know? Clin Cardiol. 2011;34:478-482.
- Khoury H, Lavoie L, Welner S, Folkerts K. The burden of major adverse cardiac events and antiplatelet prevention in patients with coronary or peripheral arterial disease. Cardiovasc Ther. 2016;34:115-124.
- Berger A, Simpson A, Bhagnani T, et al. Incidence and cost of major adverse cardiovascular events and major adverse limb events in patients with chronic coronary artery disease or peripheral artery disease. Am J Cardiol. 2019;123:1893-1899.
- American Heart Association. Cardiovascular disease: a costly burden for America: projections through 2035. Washington, DC: 2017. https://healthmetrics.heart.org/wp-content/uploads/2017/10/Cardiovascular-Disease-A-Costly-Burden.pdf. Accessed October 22, 2019.
- Mahoney EM, Wang K, Cohen DJ, et al; for the REACH Registry investigators. One-year costs in patients with a history of or at risk for atherothrombosis in the United States. Circ Cardiovasc Qual Outcomes. 2008;1:38-45.
- Korsnes JS, Davis KL, Ariely R, et al. Health care resource utilization and costs associated with nonfatal major adverse cardiovascular events. J Manag Care Spec Pharm. 2015;21:443-450.
- Eikelboom JW, Connolly SJ, Bosch J, et al; for the COMPASS investigators. Rivaroxaban with or without aspirin in stable cardiovascular disease. N Engl J Med. 2017;377:1319-1330.
- Anand SS, Bosch J, Eikelboom JW, et al; for the COMPASS investigators. Rivaroxaban with or without aspirin in patients with stable peripheral or carotid artery disease: an international, randomised, double-blind, placebo-controlled trial. Lancet. 2018;391:219-229.
- Connolly SJ, Eikelboom JW, Bosch J, et al; for the COMPASS investigators. Rivaroxaban with or without aspirin in patients with stable coronary artery disease: an international, randomised, double-blind, placebo-controlled trial. Lancet. 2018;391:205-218. Erratum in: Lancet. 2018;391:204.
- Cowie MR, Lamy A, Levy P, et al. Health economic evaluation of rivaroxaban in the treatment of patients with chronic coronary artery disease or peripheral artery disease. Cardiovasc Res. 2020;116:1918-1924.
- IBM. IBM Micromedex RED BOOK online. www.ibm.com/us-en/marketplace/micromedex-red-book. Accessed October 22, 2019. [Requires subscription to access.]
- Cunningham A, Stein CM, Chung CP, et al. An automated database case definition for serious bleeding related to oral anticoagulant use. Pharmacoepidemiol Drug Saf. 2011;20:560-566.
- U.S. Bureau of Labor Statistics. Consumer Price Index. www.bls.gov/cpi/. Accessed September 16, 2019.
- Anand SS, Eikelboom JW, Dyal L, et al; for the COMPASS trial investigators. Rivaroxaban plus aspirin versus aspirin in relation to vascular risk in the COMPASS trial. J Am Coll Cardiol. 2019;73:3271-3280.
- Mauskopf J, Earnshaw S. A methodological review of US budget-impact models for new drugs. Pharmacoeconomics. 2016;34:1111-1131.