- Email: [email protected]
Cost-Effectiveness of Drug-Eluting Stents Including the Economic Impact of Late Stent Thrombosis
Cost-Effectiveness of Drug-Eluting Stents Including the Economic Impact of Late Stent Thrombosis Kristian B. Filion, MSca,b,c, Arup Michael Roy, MDa, Tara Baboushkin, BAa, Stéphane Rinfret, MD, MScd, and Mark J. Eisenberg, MD, MPHa,b,* Recent studies examining the effectiveness of drug-eluting stents (DES) have found that the use of DES is associated with a significant increase in the incidence of late stent thrombosis (LST). Previous cost-effectiveness analyses of DES have not accounted for the costs associated with LST. In this study, published research was reviewed to identify studies that compared the cost-effectiveness of DES with that of bare-metal stents and to identify the incidence of LST. Probable costs were assigned to LST-related myocardial infarction and death on the basis of the treatment costs for these outcomes. These costs as well as those of extended clopidogrel therapy were then incorporated into the Sirolimus-Eluting Balloon Expandable Stent in the Treatment of Patients With De Novo Native Coronary Artery Lesions (SIRIUS) and TAXUS-IV cost-effectiveness data. This review found that the incidence of LST ranged from 0.2% to 0.7%. Assuming a base case LST incidence of 0.5%, a cost per death of $20,000, a cost per myocardial infarction of $20,000, and a cost of an additional 2 years of clopidogrel therapy of $2,428 per patient, the costs per revascularization avoided were $15,056 for the SIRIUS trial and $25,210 for the TAXUS-IV trial. The costs per quality-adjusted life-year gained were $250,935 and $257,591, respectively. Sensitivity analyses revealed that the costs per revascularization avoided varied from $14,618 to $15,830 for the SIRIUS trial and from $24,540 to $26,396 for the TAXUS-IV trial. Similarly, the cost per quality-adjusted life-year gained varied from $243,638 to $263,840 for the SIRIUS trial and from $250,739 to $269,708 for the TAXUS-IV trial. In conclusion, LST-related adverse events and the need for extended clopidogrel therapy substantially increase the costs associated with the implementation of DES. The inclusion of these costs renders the widespread use of DES not cost effective in the United States in terms of cost per quality-adjusted life-year gained and cost per revascularization avoided. © 2009 Elsevier Inc. (Am J Cardiol 2009;103:338 –344)
Recent studies examining the effectiveness of drug-eluting stents (DES) have found that the use of DES is associated with a significant increase in the incidence of late stent thrombosis (LST) and very late stent thrombosis.1– 4 LST, beginning 31 days after implantation, is responsible for a small but important increase in myocardial infarction and death in DES recipients, and this increase may negate the previously reported benefits associated with the use of DES. These benefits include decreased restenosis and repeat procedures but no effect on mortality or reinfarction.5 Previous studies have examined the cost-effectiveness of DES and have yielded varying results.6 –18 However, these studies a
Divisions of Cardiology and Clinical Epidemiology, Jewish General Hospital/McGill University; bDepartment of Epidemiology, Biostatistics, and Occupational Health, McGill University; cDivision of Clinical Epidemiology, McGill University Health Center, Montreal, Quebec; and dCentre de Recherche de l’Hôpital Laval, Quebec, Quebec, Canada. Dr. Eisenberg is a Chercheur-National of the Fonds de la Recherche en Santé du Québec. Dr. Rinfret is a Junior Physician-Scientist of the Fonds de la Recherche en Santé du Québec. Mr. Filion is supported, in part, by a bursary from the Fonds de la Recherche en Santé du Québec. Manuscript received June 18, 2008; revised manuscript received and accepted September 25, 2008. *Corresponding author: Tel: 514-340-8222 ext. 3564; fax: 514-3407564. E-mail address: [email protected] (M.J. Eisenberg). 0002-9149/09/$ – see front matter © 2009 Elsevier Inc. doi:10.1016/j.amjcard.2008.09.086
have been limited by an average follow-up period of only 1 year and thus have not incorporated the costs associated with the occurrence of LST and its related adverse events. These studies also have not considered the cost of extended clopidogrel therapy, which has been recommended because of the occurrence of LST.19 Consequently, we examined the effect of LST on the cost-effectiveness of DES. Methods To date, 13 cost-effectiveness analyses of DES have been conducted, 7 in North America and 6 in Europe and Australia.6 –18 Two of these analyses were prospective trials examining the cost-effectiveness of DES compared with bare-metal stents (BMS) in the United States. The first was conducted as a substudy of the Sirolimus-Eluting Balloon Expandable Stent in the Treatment of Patients With De Novo Native Coronary Artery Lesions (SIRIUS) trial,13 which compared sirolimus-eluting stents with BMS, and the second was a substudy of the TAXUS-IV trial,14 which compared paclitaxel-eluting stents with BMS. In the present study, we extrapolated the results of the SIRIUS and TAXUS-IV trials by incorporating the anticipated costs of adverse events due to LST. We selected these cost-effectiveness studies because they were the only studies that www.AJConline.org
Coronary Artery Disease/Economic Impact of Late Stent Thrombosis
339
Table 1 Baseline clinical and angiographic characteristics of patients in the Sirolimus-Eluting Balloon Expandable Stent in the Treatment of Patients With De Novo Native Coronary Artery Lesions and TAXUS-IV trials SIRIUS13,ⴱ
Characteristic
Age (yrs) Men Diabetes mellitus Current smokers Previous myocardial infarction Multivessel coronary disease Ejection fraction (%) Coronary narrowing location Left anterior descending Circumflex Right Lesion length (mm) Reference diameter (mm)
TAXUS-IV14
Sirolimus-Eluting Stents (n ⫽ 533)
BMS (n ⫽ 522)
Paclitaxel-Eluting Stents (n ⫽ 662)
BMS (n ⫽ 652)
62 ⫾ 11 72.6% 24.6% 17.7% 28.2% 40.7% 56 ⫾ 10
62 ⫾ 11 69.7% 28.2% 22.4% 32.9% 42.5% 56 ⫾ 10
63 ⫾ 11 71.8% 23.4% 23.4% 30.5% — 55 ⫾ 10
62 ⫾ 11 72.4% 25.0% 20.1% 29.9% — 55 ⫾ 10
44.1% 25.2% 30.1% 14.4 ⫾ 5.8 2.79 ⫾ 0.45
43.2% 23.9% 32.4% 14.4 ⫾ 45.8 2.81 ⫾ 0.49
40.0% 28.9% 31.1% 13.4 ⫾ 6.3 2.75 ⫾ 0.47
41.4% 26.6% 32.0% 13.4 ⫾ 6.2 2.75 ⫾ 0.49
Data are expressed as mean ⫾ SD or as percentages. For all comparisons, p ⫽ NS. Adapted with permission from Circulation13 and J Am Coll Cardiol.14 * In the SIRIUS trial, 1,058 patients were randomized. However, only 1,055 patients were included in the corresponding cost-effectiveness study.
prospectively collected cost data in a North American context. In the 2 studies, these data were collected meticulously. The primary outcome measures used in previous DES cost-effectiveness analyses were the incremental cost-effectiveness ratio (ICER), measured in cost per quality-adjusted life-year (QALY) gained, and cost per revascularization avoided.20,21 The ICER is generally calculated by determining the difference in cost between 2 treatment alternatives and dividing the difference in cost by the difference in QALYs,22 a measure that accounts for the quantity and quality of life gained.23 In the United States, an ICER ⬍$50,000 per QALY gained is deemed economically favorable, an ICER of $50,000 to $100,000 per QALY gained is considered to be in the “gray zone,” and an ICER ⬎$100,000 per QALY gained is not attractive.12 Although QALYs are a widely accepted and commonly used measure of cost-effectiveness, their use in assessing the value of treatments intended to prevent restenosis is limited. This limited value is due to the lack of evidence indicating that restenosis affects survival. Therefore, in patients who undergo percutaneous coronary intervention, QALYs will depend primarily on the methods (quality and intensity) of the measure of quality of life, or utility scores, not the occurrence of clinical events. Consequently, the treatment of restenosis will have only a marginal effect on quality-adjusted life expectancy. To overcome the limitations of QALYs in this setting, recent studies have applied a cost-effective ratio that is disease specific to patients who undergo percutaneous coronary intervention: cost per revascularization avoided. In the United States, a cost ⬍$10,000 per repeat revascularization avoid is considered to be costeffective.12 The cost per revascularization avoided is traditionally obtained by dividing the difference in 1-year medical costs by the difference in the number of revascularization procedures over that time, and the $10,000 threshold is based on the amounts reimbursed by most third-party payers.12,24
In the SIRIUS trial, 1,058 patients with moderately complex coronary stenoses were randomized to percutaneous coronary intervention with either sirolimus-eluting stents or BMS.13 Clinical outcomes, resource use, and costs were assessed prospectively for all patients over a follow-up period of 1 year. Inclusion criteria included a de novo lesion measuring 15 to 30 mm in length and 2.5 to 3.5 mm in width in a native coronary artery. All patients received aspirin and clopidogrel after percutaneous coronary intervention for ⱖ3 months. The acquisition costs of sirolimus-eluting stents and BMS were estimated at $2,900 and $900 per stent, respectively. Baseline demographic and clinical characteristics of the patients in the SIRIUS trial are listed in Table 1. The cost-effectiveness analyses of the SIRIUS trial included the resource use and costs of the index revascularization procedure; the in-hospital adverse events, resource use, and costs of the index hospitalization; and adverse events, resource use, and costs accrued up to 1 year, including the cost per repeat revascularization (Table 2). In the SIRIUS trial, patients randomized to DES had costs of index hospitalization that were $2,881 higher than patients randomized to BMS (Table 2).13 However, patients randomized to BMS had total follow-up costs that were higher than patients randomized to DES. Consequently, the aggregate 1-year costs were similar between groups. In the TAXUS-IV trial, 1,314 patients were randomized to percutaneous coronary intervention with either paclitaxeleluting stents or BMS.14 As in SIRIUS, clinical outcomes, resource use, and costs were assessed prospectively over a 1-year period. Inclusion criteria included a de novo lesion measuring 10 to 28 mm in length and 2.5 to 3.75 mm in width in a native coronary artery. All patients received antiplatelet therapy for ⱖ6 months after percutaneous coronary intervention. The acquisition costs of paclitaxel-eluting stents and BMS were estimated at $2,700 and $800 per stent, respectively. Baseline characteristics are listed in Table 1. Resource use and costs of the index revascularization
340
The American Journal of Cardiology (www.AJConline.org)
Table 2 Index procedure, in-hospital, and follow-up resource use and costs for patients in the Sirolimus-Eluting Balloon Expandable Stent in the Treatment of Patients With De Novo Native Coronary Artery Lesions and TAXUS-IV trials SIRIUS13,†
Characteristic
Sirolimus-Eluting Stents (n ⫽ 533) Procedural characteristics Procedure duration (min) Contrast volume (ml) Guiding catheters Guidewires Balloon catheters Stents Resource costs ($) Room/overhead Supplies/drugs Devices Personnel* Total procedural cost ($) In-hospital characteristics Death Myocardial infarction Repeat percutaneous coronary intervention Coronary artery bypass graft surgery Diagnostic catheterization Vascular complication/transfusion Length of stay (d) Postprocedural length of stay (d) Medical costs ($) Initial procedure Hospital room/ancillary Professional fees Total Follow-up characteristics Death Myocardial infarction Repeat revascularization Coronary artery bypass graft surgery Percutaneous coronary intervention Repeat hospitalization Revascularization procedures (per 100 patients) Coronary artery bypass graft surgery Percutaneous coronary intervention Diagnostic catheterizations (per 100 patients) Hospital admissions (per 100 patients) Hospital days (per 100 patients) Follow-up costs ($) Repeat procedures Hospital room/ancillary Inpatient physician fees Outpatient services/physician fees Total follow-up costs Aggregate 1-yr costs ($)
76 ⫾ 35 220 ⫾ 112 1.3 ⫾ 0.6 1.4 ⫾ 0.8 1.5 ⫾ 0.7 1.4 ⫾ 0.8 1,222 ⫾ 313 (1,190) 1,026 ⫾ 582 (939) 4,865 ⫾ 1,936 (3,700) 131 ⫾ 35 (126) 7,251 ⫾ 2,142 (6,643) 0.2% 2.3% 0.2% 0.2% 0.4% 1.7% 1.7 ⫾ 1.4 (1) 1.2 ⫾ 0.9 (1)
TAXUS-IV14 BMS (n ⫽ 522)
77 ⫾ 39 204 ⫾ 105 1.3 ⫾ 0.7 1.3 ⫾ 0.7 1.5 ⫾ 0.8 1.4 ⫾ 0.6 1,237 ⫾ 529 (1,163) 975 ⫾ 557 (929) 2,052 ⫾ 738 (1,700) 131 ⫾ 59 (123) 4,395 ⫾ 1,263 (4,302) 0.0% 1.5% 0.0% 0.0% 0.2% 2.3% 1.7 ⫾ 1.2 (1) 1.2 ⫾ 0.6 (1)
Paclitaxel-Eluting Stents (n ⫽ 662) 51 ⫾ 27 — — — 2.9 ⫾ 2.4 1.3 ⫾ 0.7
BMS (n ⫽ 652) 51 ⫾ 28 — — — 2.8 ⫾ 2.4 1.3 ⫾ 0.8
— — 3,966 ⫾ 1,363 (3,380)
— — 1,924 ⫾ 1,193 (1,440)
6,324 ⫾ 2,188 (5,740)
4,336 ⫾ 2,427 (3,852)
0.0% 2.4% 0.3% 0.0% 1.2% 1.6% 2.0 ⫾ 2.0 (1) 1.3 ⫾ 0.9 (1)
0.0% 2.1% 0.2% 0.2% 0.3% 1.9% 1.9 ⫾ 2.1 (1) 1.3 ⫾ 1.6 (1)
7,251 ⫾ 2,142 (6,643) 2,701 ⫾ 1,997 (1,832) 1,379 ⫾ 367 (1,284) 11,345 ⫾ 3,211 (10,462)
4,395 ⫾ 1,263 (4,302) 2,704 ⫾ 1,865 (1,832) 1,362 ⫾ 193 (1,284) 8,464 ⫾ 2,497 (7,884)
6,324 ⫾ 2,188 (5,740) 2,882 ⫾ 1,858 (2,497) 1,889 ⫾ 340 (1,749) 11,096 ⫾ 3,195 (10,165)
4,336 ⫾ 2,427 (3,852) 2,849 ⫾ 1,960 (2,497) 1,883 ⫾ 586 (1,749) 9,067 ⫾ 3,387 (8,230)
1.1% 0.8% 13.3% 1.3% 12.4% 27.0% 16.3 1.3 15.0 12.8 40.3 109.0
0.8% 1.9% 28.4% 3.0% 26.9% 41.1% 35.4 3.0 32.4 18.5 64.8 148.2
2.1% 1.1% 6.6% 1.7% 5.1% 18.6% 6.9 1.7 5.3 — 26.4 71.3
1.4% 2.5% 16.6% 3.8% 13.3% 26.4% 19.2 3.8 15.3 — 38.0 104.9
— — — — 3,487 (540) 14,583 (11,699)
— — — — 4,944 (90) 14,011 (9,540)
789 1,936 1,280 1,463 5,468 16,813
(0) (0) (0) (444) (2,522) (14,000)
1,788 3,149 1,759 1,343 8,040 16,504
(0) (0) (0) (728) (4,125) (12,802)
Data are expressed as mean ⫾ SD, numbers, or percentages. Values in parentheses are medians. Adapted with permission from Circulation13 and J Am Coll Cardiol.14 * Personnel costs include only nonphysician personnel. † In the SIRIUS trial, 1,058 patients were randomized. However, only 1,055 patients were included in the corresponding cost-effectiveness study.
procedure and adverse events, resource use, and costs of the index hospitalization and accrued up to 1 year, including the cost per repeat revascularization, are listed in Table 2. As in the SIRIUS trial, randomization to DES was associated with higher in-hospital costs during index hospitalization, whereas
BMS were associated with higher 1-year follow-up costs, resulting in similar aggregate costs at 1 year. To identify the incidence of LST (including very late stent thrombosis) in patients using DES, we conducted a systematic search of PubMed (National Library of Medi-
Coronary Artery Disease/Economic Impact of Late Stent Thrombosis Table 3 The effect of the cost of adverse events related to late stent thrombosis on the cost-effectiveness of drug-eluting stents Variable Incidence of LST (%) LST death (%) LST nonfatal myocardial infarction (%) Cost per death ($) Cost per nonfatal myocardial infarction ($) Cost of 2 additional years of clopidogrel ($) Cost per revascularization avoided ($) SIRIUS TAXUS-IV Cost per QALY gained ($) SIRIUS TAXUS-IV
Best-Case Scenario
Likely Scenario
0.2 40 60
0.5 40 60
Worst-Case Scenario 0.7 40 60
0 15,000
20,000 20,000
50,000 25,000
2,429
2,429
2,429
14,618 24,540
15,056 25,210
15,830 26,396
243,638 250,739
250,935 257,591
263,840 269,708
cine, Bethesda, Maryland) using the key words “DES” and “stent thrombosis.” We restricted our search to articles in English presenting pooled analyses comparing the incidence of LST in patients with DES to that in patients with BMS. We also restricted our search to studies that included ⱖ1,500 patients. Our search identified 5 studies, which reported increases in the incidence of LST that ranged from 0.2% to 0.7%.3,4,25,26,31 Therefore, we estimated the increase in the incidence of LST to be the middle of the range (0.5%) in patients receiving DES compared with patients receiving BMS, and we varied this value to the extremes of the reported range in sensitivity analyses. We also estimated that 40% of all episodes of LST were fatal, with the remaining 60% consisting of nonfatal myocardial infarctions. This case fatality rate is consistent with that observed in the SIRIUS trial. We calculated the ICER, expressed in cost per QALY gained, and the cost per revascularization avoided from the perspective of a third-party payer by incorporating the costs associated with LST into the cost-effectiveness analyses of the SIRIUS and TAXUS-IV trials. The costs associated with LST included the cost of nonfatal myocardial infarction, which we estimated to be $20,000 per nonfatal myocardial infarction, which is the approximate cost of repeat percutaneous coronary intervention,27,28 and the costs of death. To estimate the cost of death, we assumed that 50% of patients died suddenly with no treatment costs and that 50% of patients died after the initiation of treatment. We assumed that patients who died after the initiation of treatment had substantially higher treatment costs than those with nonfatal myocardial infarctions. Consequently, we assigned an average cost of $20,000 per death. In light of the new treatment guidelines recommending extended clopidogrel therapy in patients with DES,19 we also incorporated 2 additional years of clopidogrel therapy at a cost of $2,428 per DES patient.29 These costs were added to the cost per QALY gained and cost per revascularization originally presented in the SIRIUS and TAXUS-IV trials to obtain the
341
trial-specific relative increase in cost per patient. We then used these relative increases to obtain an ICER and cost per revascularization avoided that included the costs associated with LST. In sensitivity analyses, we examined the bestand worst-case scenarios by varying the increased LST incidence from 0.2% to 0.7%, the cost per death from $0 to $50,000, and the cost per myocardial infarction from $15,000 to $25,000. All costs are presented in United States dollars. Results Using an incidence of LST of 0.5% and a case fatality rate of 40%,26 2 additional deaths and 3 additional myocardial infarctions per 1,000 patients treated with DES are expected compared with the number of events originally reported in DES trials. Assuming a corresponding cost of $20,000 per death and $20,000 per myocardial infarction, and incorporating the cost of extended clopidogrel therapy, these additional events are associated with an increased cost of $2,510,600 per 1,000 patients treated with DES. The inclusion of these LST-associated costs results in a relative increase in cost per patient of 912% in the SIRIUS trial and 539% in the TAXUS-IV trial (Table 3). The costs associated with LST have important implications for measures of the cost-effectiveness of DES (Table 3). When these costs are taken into consideration, the cost per revascularization avoided increased from $1,650 in the SIRIUS trial to $15,056 and from $4,678 in the TAXUS-IV trial to $25,210. Similarly, LST-associated costs increased the cost per QALY gained from $27,500 in the SIRIUS trial to $250,935 and from $47,798 in the TAXUS-IV trial to $257,591. To examine the robustness of our estimates, we conducted sensitivity analyses in which we varied key parameters to produce best-case and worst-case scenarios (Table 3). This included varying the (1) the incidence of LST from 0.2% to 0.7%, (2) the cost per death from $0 to $50,000, and (3) the cost per myocardial infarction from $15,000 to $25,000. The effects of varying the incidence of LST from 0.2% to 0.7% are depicted in Figures 1 and 2. Under the assumptions for a best-case scenario, the cost per revascularization avoided was $14,618 in the SIRIUS trial and $24,540 in the TAXUS-IV trial. The cost per QALY gained in the best-case scenario was $243,638 in the SIRIUS trial and $250,739 in the TAXUS-IV trial. Under the assumptions for a worst-case scenario, the cost per revascularization avoided was $15,830 in the SIRIUS trial and $26,396 in the TAXUS-IV trial. The cost per QALY gained in the worst-case scenario was $263,840 in the SIRIUS trial and $269,708 in the TAXUS-IV trial. Discussion Our study was designed to examine the effect of LST on the cost-effectiveness of DES. Using data from the SIRIUS and TAXUS-IV trials, we found that LST, although rare, has a substantial effect on the cost-effectiveness of DES. The costs associated with LST increased the cost per revascularization avoided to $15,056 using data from the SIRIUS trial and to $25,210 using data from the TAXUS-IV trial.
342
The American Journal of Cardiology (www.AJConline.org)
Figure 1. Effect of varying the incidence of LST on the ICER in cost per QALY gained. The incidence of LST was varied from 0.2% to 0.7%, representing the range identified in a review of the published research. These sensitivity analyses are based on a cost of $20,000 per death, $20,000 per myocardial infarction, and 2 additional years of clopidogrel therapy in patients with DES. The dashed lines represent the accepted threshold for cost-effectiveness of $50,000 per QALY and the threshold for the gray zone of cost-effectiveness of $100,000 per QALY.
Similarly, LST-associated costs increased the ICER to $250,935 and $257,591 per QALY gained, respectively. Consequently, when the LST-associated costs are incorporated, DES exceeds the accepted thresholds of $100,000 per QALY (the limit of the gray zone of cost-effectiveness) and $10,000 per revascularization avoided, and thus, DES is no longer cost-effective. This finding was present using data from the SIRIUS and TAXUS-IV trials and for the 2 measures of cost-effectiveness. The costs associated with LST and corresponding increases in the ICER and cost per revascularization avoided have important policy implications. Previous studies have suggested that DES, although not cost effective from a hospital perspective, are cost effective from a societal perspective.21 However, the costs associated with LST shift measures of cost-effectiveness beyond the accepted thresholds. With limited resources available for health care and the advent of expensive treatment options (e.g., internal cardiac defibrillators, left ventricular assist devices) that are marginally cost effective, these findings confirm that the widespread use of DES is not justifiable.20 These findings are particularly important for the United States, where until recently, DES were used in ⬎80% of percutaneous coronary intervention procedures.30 Importantly, with no difference in mortality, previous cost-effectiveness analyses were based almost entirely on the observed difference in quality of life. Consequently, these analyses involved trading cost for improved quality of life in the absence of LST. However, once LST is incorpo-
rated, cost-effectiveness analyses now involve trading cost for improved quality of life and small but important increases in death and nonfatal myocardial infarction. The controversy surrounding DES has led to many costeffectiveness analyses.6 –18 In the United States, only 2 prospective cost-effectiveness analyses have been conducted.13,14 In these 2 studies, the investigators concluded that DES are cost effective in high-risk patients, on the basis of the cost per revascularization avoided. Cohen et al13 also reported an ICER in SIRIUS that is considered to be cost effective in the United States. The cost-effectiveness of DES has also been examined in 5 studies from Canada,6,7,15–17 as well as studies from Australia,8 Sweden,9 Switzerland,10 and the United Kingdom.11,18 All 5 Canadian studies found that DES were far from cost effective when used in a widespread, indiscriminate manner and may be only borderline cost effective in a very select percutaneous coronary intervention patient subgroup at the highest risk for restenosis. Cost-effectiveness analyses from Europe and Australia provided results that were similar to those reported in Canadian studies. These studies include the BAsel Stent Kosten Effektivitäts Trial (BASKET), which attempted to reproduce a “real-world” scenario by not implementing protocol-mandated follow-up angiography.10 The BASKET investigators concluded that in a real-world setting, DES should be reserved for patients in high-risk groups. The investigators also raised concerns regarding thrombotic complications associated with DES in the 6 months after the cessation of clopidogrel. In light of
Coronary Artery Disease/Economic Impact of Late Stent Thrombosis
343
Figure 2. Effect of varying the incidence of LST on the cost per revascularization avoided. The incidence of LST was varied from 0.2% to 0.7%, representing the range identified in a review of the published research. These sensitivity analyses are based on a cost of $20,000 per death, $20,000 per MI, and 2 additional years of clopidogrel therapy in patients with DES. The dashed line represents the accepted threshold for cost-effectiveness of $10,000 per revascularization avoided.
their concerns regarding thrombotic complications, the investigators suggested a need for prolonged clopidogrel therapy. Importantly, as with the other cost-effectiveness analyses of DES, the BASKET trial did not consider the costs of LST or the associated need for extended clopidogrel therapy. These additional costs render DES not cost effective in the United States, the only setting to date where the widespread use of DES may have been cost effective. The inclusion of these costs in cost-effectiveness analyses from other countries would further highlight the unfavorable cost-effectiveness of the widespread use of DES. Our cost-effectiveness analysis has a number of potential limitations. First, we only examined costs from a third-party payer perspective and did not included the societal costs associated with the loss of productivity or the humanistic cost associated with the additional loss of life. As a result, we underestimated the costs associated with LST, and thus, our ICER and cost per revascularization avoided represent conservative estimates. Second, we used conservative assumptions (e.g., incidence of 0.5%, 40% mortality) and conservative LST-associated costs (e.g., $20,000 per myocardial infarction, $20,000 per death). We also assumed that all repeat revascularizations were percutaneous coronary interventions rather than bypass surgeries, which would increase the cost of treating LST-associated events. It is therefore likely that we underestimated the costs associated with LST. Third, we also only included the cost of an additional 2 years of clopidogrel therapy. The inclusion of the cost associated with indefinite dual-antiplatelet therapy
would further increase the ICER and cost per revascularization avoided of DES. Fourth, it is possible that sustained clopidogrel therapy would reduce the observed rate of LST. However, in the extreme scenario of an LST incidence rate of 0%, the cost of clopidogrel alone renders DES not cost effective. Our analysis is based on aggregate published data from the SIRIUS and TAXUS-IV trials. Consequently, we were unable to examine the cost-effectiveness of DES among various subgroups of patients. Thus, although the widespread use of DES is no longer cost effective, their costeffectiveness in high-risk patients remains unknown. In addition, as is true with most trials, the generalizability of their results is unclear. Our analysis is also limited by the limitations of the original SIRIUS and TAXUS-IV costeffectiveness analyses. These limitations include minimizing the number of stents used per procedure and the use of protocol-mandated angiography. Minimizing the number of stents used per percutaneous coronary intervention results in a lower cost per repeat revascularization avoided and consequently a more favorable cost-effectiveness ratio. The inclusion of protocol-mandated angiography results in the reporting of higher rates of restenosis and repeat revascularization procedures in BMS patients and ultimately a more favorable estimate of the clinical effectiveness, and subsequently cost-effectiveness, of DES. These limitations of the SIRIUS and TAXUS-IV cost-effectiveness studies likely underestimate the true ICER and cost per revascularization
344
The American Journal of Cardiology (www.AJConline.org)
avoided of DES, exaggerating the cost-effectiveness of DES. 1. Bavry AA, Kumbhani DJ, Helton TJ, Borek PP, Mood GR, Bhatt DL. Late thrombosis of drug-eluting stents: a meta-analysis of randomized clinical trials. Am J Med 2006;119:1056 –1061. 2. Iakovou I, Schmidt T, Bonizzoni E, Ge L, Sangiorgi GM, Stankovic G, Airoldi F, Chieffo A, Montorfano M, Carlino M, et al. Incidence, predictors, and outcome of thrombosis after successful implantation of drug-eluting stents. JAMA 2005;293:2126 –2130. 3. Lagerqvist B, James SK, Stenestrand U, Lindback J, Nilsson T, Wallentin L. Long-term outcomes with drug-eluting stents versus baremetal stents in Sweden. N Engl J Med 2007;356:1009 –1019. 4. Stone GW, Moses JW, Ellis SG, Schofer J, Dawkins KD, Morice MC, Colombo A, Schampaert E, Grube E, Kirtane AJ, et al. Safety and efficacy of sirolimus- and paclitaxel-eluting coronary stents. N Engl J Med 2007;356:998 –1008. 5. Babapulle MN, Joseph L, Belisle P, Brophy JM, Eisenberg MJ. A hierarchical Bayesian meta-analysis of randomised clinical trials of drug-eluting stents. Lancet 2004;364:583–591. 6. Shrive FM, Manns BJ, Galbraith PD, Knudtson ML, Ghali WA. Economic evaluation of sirolimus-eluting stents. CMAJ 2005;172: 345–351. 7. Rinfret S, Cohen DJ, Tahami Monfared AA, Lelorier J, Mireault J, Schampaert E. Cost effectiveness of the sirolimus-eluting stent in high-risk patients in Canada: an analysis from the C-SIRIUS trial. Am J Cardiovasc Drugs 2006;6:159 –168. 8. Lord SJ, Howard K, Allen F, Marinovich L, Burgess DC, King R, Atherton JJ. A systematic review and economic analysis of drug-eluting coronary stents available in Australia. Med J Aust 2005;183:464 – 471. 9. Ekman M, Sjogren I, James S. Cost-effectiveness of the Taxus paclitaxel-eluting stent in the Swedish healthcare system. Scand Cardiovasc J 2006;40:17–24. 10. Kaiser C, Brunner-La Rocca HP, Buser PT, Bonetti PO, Osswald S, Linka A, Bernheim A, Zutter A, Zellweger M, Grize L, Pfisterer ME. Incremental cost-effectiveness of drug-eluting stents compared with a third-generation bare-metal stent in a real-world setting: randomised Basel Stent Kosten Effektivitats Trial (BASKET). Lancet 2005;366: 921–929. 11. Bagust A, Grayson AD, Palmer ND, Perry RA, Walley T. Cost effectiveness of drug eluting coronary artery stenting in a UK setting: cost-utility study. Heart 2006;92:68 –74. 12. Greenberg D, Bakhai A, Cohen DJ. Can we afford to eliminate restenosis? Can we afford not to? J Am Coll Cardiol 2004;43:513–518. 13. Cohen DJ, Bakhai A, Shi C, Githiora L, Lavelle T, Berezin RH, Leon MB, Moses JW, Carrozza JP Jr, Zidar JP, Kuntz RE. Cost-effectiveness of sirolimus-eluting stents for treatment of complex coronary stenoses: results from the Sirolimus-Eluting Balloon Expandable Stent in the Treatment of Patients With De Novo Native Coronary Artery Lesions (SIRIUS) trial. Circulation 2004;110:508 –514. 14. Bakhai A, Stone GW, Mahoney E, Lavelle TA, Shi C, Berezin RH, Lahue BJ, Clark MA, Lacey MJ, Russell ME, et al. Cost effectiveness of paclitaxel-eluting stents for patients undergoing percutaneous coronary revascularization: results from the TAXUS-IV trial. J Am Coll Cardiol 2006;48:253–261. 15. Bowen J, Hopkins R, He Y, Blackhouse G, Lazzam C, Tu JV, Cohen E, Tarride JE, Goeree R. Systematic Review and Cost-Effectiveness
16.
17.
18.
19.
20. 21. 22.
23. 24.
25.
26.
27.
28. 29.
30. 31.
Analysis of Drug Eluting Stents Compared to Bare Metal Stents for Percutaneous Coronary Interventions in Ontario: Interim Report for the Ontario Ministry of Health and Long-Term Care. Report No. HTA002– 0512. Hamilton, Ontario, Canada: Ontario Ministry of Health and Long-Term Care, 2005. Mittmann N, Brown A, Seung SJ, Coyle D, Cohen E, Brophy J, Title L, Oh P. Economic Evaluation of Drug Eluting Stents. Report No. 53 Ottawa, Ontario, Canada: Canadian Coordinating Office for Health Technology Assessment, 2005. Brophy J, Erickson L. An Economic Analysis of Drug Eluting Coronary Stents: A Quebec Perspective. Report No. 04-04. Montreal, Quebec, Canada: Agence d’Evaluation des Technologies et des Modes d’Intervention en Sante, 2004. Liverpool Reviews & Implementation Group. Coronary Artery Stents Rapid Systematic Review and Economic Evaluation (Addendum B). Commissioned by NHS R&D HTA Programme on Behalf of the National Institute for Clinical Excellence. Liverpool, United Kingdom: Liverpool Reviews & Implementation Group, 2003. King SB III, Smith SC Jr, Hirshfeld JW Jr, Jacobs AK, Morrison DA, Williams DO, Feldman TE, Kern MJ, O’Neill WW, Schaff HV, et al. 2007 focused update of the ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2008;51:172–209. Eisenberg MJ. Drug-eluting stents: the price is not right. Circulation 2006;114:1745–1754. Ryan J, Cohen DJ. Are drug-eluting stents cost-effective? It depends on whom you ask. Circulation 2006;114:1736 –1743. Gaziano TA, Opie LH, Weinstein MC. Cardiovascular disease prevention with a multidrug regimen in the developing world: a cost-effectiveness analysis. Lancet 2006;368:679 – 686. Weinstein MC, Stason WB. Foundations of cost-effectiveness analysis for health and medical practices. N Engl J Med 1977;296:716 –721. Cohen DJ, Taira DA, Berezin R, Cox DA, Morice MC, Stone GW, Grines CL. Cost-effectiveness of coronary stenting in acute myocardial infarction: results from the Stent Primary Angioplasty in Myocardial Infarction (Stent-PAMI) trial. Circulation 2001;104:3039 –3045. Kastrati A, Mehilli J, Pache J, Kaiser C, Valgimigli M, Kelbaek H, Menichelli M, Sabate M, Suttorp MJ, Baumgart D, et al. Analysis of 14 trials comparing sirolimus-eluting stents with bare-metal stents. N Engl J Med 2007;356:1030 –1039. Mauri L, Hsieh WH, Massaro JM, Ho KK, D’Agostino R, Cutlip DE. Stent thrombosis in randomized clinical trials of drug-eluting stents. N Engl J Med 2007;356:1020 –1029. Yock CA, Boothroyd DB, Owens DK, Garber AM, Hlatky MA. Cost-effectiveness of bypass surgery versus stenting in patients with multivessel coronary artery disease. Am J Med 2003;115:382–389. Nagle PC, Smith AW. Review of recent US cost estimates of revascularization. Am J Manag Care 2004;10(suppl):S370 –S376. Cowper PA, Udayakumar K, Sketch MH Jr, Peterson ED. Economic effects of prolonged clopidogrel therapy after percutaneous coronary intervention. J Am Coll Cardiol 2005;45:369 –376. Maisel WH. Unanswered questions— drug-eluting stents and the risk of late thrombosis. N Engl J Med 2007;356:981–984. Spaulding C, Daemen J, Boersma E, Cutlip DE, Serruys PW. A pooled analysis of data comparing sirolimus-eluting stents with bare-metal stents. N Engl J Med 2007;356:989 –997.
Recent studies examining the effectiveness of drug-eluting stents (DES) have found that the use of DES is associated with a significant increase in the incidence of late stent thrombosis (LST) and very late stent thrombosis.1– 4 LST, beginning 31 days after implantation, is responsible for a small but important increase in myocardial infarction and death in DES recipients, and this increase may negate the previously reported benefits associated with the use of DES. These benefits include decreased restenosis and repeat procedures but no effect on mortality or reinfarction.5 Previous studies have examined the cost-effectiveness of DES and have yielded varying results.6 –18 However, these studies a
Divisions of Cardiology and Clinical Epidemiology, Jewish General Hospital/McGill University; bDepartment of Epidemiology, Biostatistics, and Occupational Health, McGill University; cDivision of Clinical Epidemiology, McGill University Health Center, Montreal, Quebec; and dCentre de Recherche de l’Hôpital Laval, Quebec, Quebec, Canada. Dr. Eisenberg is a Chercheur-National of the Fonds de la Recherche en Santé du Québec. Dr. Rinfret is a Junior Physician-Scientist of the Fonds de la Recherche en Santé du Québec. Mr. Filion is supported, in part, by a bursary from the Fonds de la Recherche en Santé du Québec. Manuscript received June 18, 2008; revised manuscript received and accepted September 25, 2008. *Corresponding author: Tel: 514-340-8222 ext. 3564; fax: 514-3407564. E-mail address: [email protected] (M.J. Eisenberg). 0002-9149/09/$ – see front matter © 2009 Elsevier Inc. doi:10.1016/j.amjcard.2008.09.086
have been limited by an average follow-up period of only 1 year and thus have not incorporated the costs associated with the occurrence of LST and its related adverse events. These studies also have not considered the cost of extended clopidogrel therapy, which has been recommended because of the occurrence of LST.19 Consequently, we examined the effect of LST on the cost-effectiveness of DES. Methods To date, 13 cost-effectiveness analyses of DES have been conducted, 7 in North America and 6 in Europe and Australia.6 –18 Two of these analyses were prospective trials examining the cost-effectiveness of DES compared with bare-metal stents (BMS) in the United States. The first was conducted as a substudy of the Sirolimus-Eluting Balloon Expandable Stent in the Treatment of Patients With De Novo Native Coronary Artery Lesions (SIRIUS) trial,13 which compared sirolimus-eluting stents with BMS, and the second was a substudy of the TAXUS-IV trial,14 which compared paclitaxel-eluting stents with BMS. In the present study, we extrapolated the results of the SIRIUS and TAXUS-IV trials by incorporating the anticipated costs of adverse events due to LST. We selected these cost-effectiveness studies because they were the only studies that www.AJConline.org
Coronary Artery Disease/Economic Impact of Late Stent Thrombosis
339
Table 1 Baseline clinical and angiographic characteristics of patients in the Sirolimus-Eluting Balloon Expandable Stent in the Treatment of Patients With De Novo Native Coronary Artery Lesions and TAXUS-IV trials SIRIUS13,ⴱ
Characteristic
Age (yrs) Men Diabetes mellitus Current smokers Previous myocardial infarction Multivessel coronary disease Ejection fraction (%) Coronary narrowing location Left anterior descending Circumflex Right Lesion length (mm) Reference diameter (mm)
TAXUS-IV14
Sirolimus-Eluting Stents (n ⫽ 533)
BMS (n ⫽ 522)
Paclitaxel-Eluting Stents (n ⫽ 662)
BMS (n ⫽ 652)
62 ⫾ 11 72.6% 24.6% 17.7% 28.2% 40.7% 56 ⫾ 10
62 ⫾ 11 69.7% 28.2% 22.4% 32.9% 42.5% 56 ⫾ 10
63 ⫾ 11 71.8% 23.4% 23.4% 30.5% — 55 ⫾ 10
62 ⫾ 11 72.4% 25.0% 20.1% 29.9% — 55 ⫾ 10
44.1% 25.2% 30.1% 14.4 ⫾ 5.8 2.79 ⫾ 0.45
43.2% 23.9% 32.4% 14.4 ⫾ 45.8 2.81 ⫾ 0.49
40.0% 28.9% 31.1% 13.4 ⫾ 6.3 2.75 ⫾ 0.47
41.4% 26.6% 32.0% 13.4 ⫾ 6.2 2.75 ⫾ 0.49
Data are expressed as mean ⫾ SD or as percentages. For all comparisons, p ⫽ NS. Adapted with permission from Circulation13 and J Am Coll Cardiol.14 * In the SIRIUS trial, 1,058 patients were randomized. However, only 1,055 patients were included in the corresponding cost-effectiveness study.
prospectively collected cost data in a North American context. In the 2 studies, these data were collected meticulously. The primary outcome measures used in previous DES cost-effectiveness analyses were the incremental cost-effectiveness ratio (ICER), measured in cost per quality-adjusted life-year (QALY) gained, and cost per revascularization avoided.20,21 The ICER is generally calculated by determining the difference in cost between 2 treatment alternatives and dividing the difference in cost by the difference in QALYs,22 a measure that accounts for the quantity and quality of life gained.23 In the United States, an ICER ⬍$50,000 per QALY gained is deemed economically favorable, an ICER of $50,000 to $100,000 per QALY gained is considered to be in the “gray zone,” and an ICER ⬎$100,000 per QALY gained is not attractive.12 Although QALYs are a widely accepted and commonly used measure of cost-effectiveness, their use in assessing the value of treatments intended to prevent restenosis is limited. This limited value is due to the lack of evidence indicating that restenosis affects survival. Therefore, in patients who undergo percutaneous coronary intervention, QALYs will depend primarily on the methods (quality and intensity) of the measure of quality of life, or utility scores, not the occurrence of clinical events. Consequently, the treatment of restenosis will have only a marginal effect on quality-adjusted life expectancy. To overcome the limitations of QALYs in this setting, recent studies have applied a cost-effective ratio that is disease specific to patients who undergo percutaneous coronary intervention: cost per revascularization avoided. In the United States, a cost ⬍$10,000 per repeat revascularization avoid is considered to be costeffective.12 The cost per revascularization avoided is traditionally obtained by dividing the difference in 1-year medical costs by the difference in the number of revascularization procedures over that time, and the $10,000 threshold is based on the amounts reimbursed by most third-party payers.12,24
In the SIRIUS trial, 1,058 patients with moderately complex coronary stenoses were randomized to percutaneous coronary intervention with either sirolimus-eluting stents or BMS.13 Clinical outcomes, resource use, and costs were assessed prospectively for all patients over a follow-up period of 1 year. Inclusion criteria included a de novo lesion measuring 15 to 30 mm in length and 2.5 to 3.5 mm in width in a native coronary artery. All patients received aspirin and clopidogrel after percutaneous coronary intervention for ⱖ3 months. The acquisition costs of sirolimus-eluting stents and BMS were estimated at $2,900 and $900 per stent, respectively. Baseline demographic and clinical characteristics of the patients in the SIRIUS trial are listed in Table 1. The cost-effectiveness analyses of the SIRIUS trial included the resource use and costs of the index revascularization procedure; the in-hospital adverse events, resource use, and costs of the index hospitalization; and adverse events, resource use, and costs accrued up to 1 year, including the cost per repeat revascularization (Table 2). In the SIRIUS trial, patients randomized to DES had costs of index hospitalization that were $2,881 higher than patients randomized to BMS (Table 2).13 However, patients randomized to BMS had total follow-up costs that were higher than patients randomized to DES. Consequently, the aggregate 1-year costs were similar between groups. In the TAXUS-IV trial, 1,314 patients were randomized to percutaneous coronary intervention with either paclitaxeleluting stents or BMS.14 As in SIRIUS, clinical outcomes, resource use, and costs were assessed prospectively over a 1-year period. Inclusion criteria included a de novo lesion measuring 10 to 28 mm in length and 2.5 to 3.75 mm in width in a native coronary artery. All patients received antiplatelet therapy for ⱖ6 months after percutaneous coronary intervention. The acquisition costs of paclitaxel-eluting stents and BMS were estimated at $2,700 and $800 per stent, respectively. Baseline characteristics are listed in Table 1. Resource use and costs of the index revascularization
340
The American Journal of Cardiology (www.AJConline.org)
Table 2 Index procedure, in-hospital, and follow-up resource use and costs for patients in the Sirolimus-Eluting Balloon Expandable Stent in the Treatment of Patients With De Novo Native Coronary Artery Lesions and TAXUS-IV trials SIRIUS13,†
Characteristic
Sirolimus-Eluting Stents (n ⫽ 533) Procedural characteristics Procedure duration (min) Contrast volume (ml) Guiding catheters Guidewires Balloon catheters Stents Resource costs ($) Room/overhead Supplies/drugs Devices Personnel* Total procedural cost ($) In-hospital characteristics Death Myocardial infarction Repeat percutaneous coronary intervention Coronary artery bypass graft surgery Diagnostic catheterization Vascular complication/transfusion Length of stay (d) Postprocedural length of stay (d) Medical costs ($) Initial procedure Hospital room/ancillary Professional fees Total Follow-up characteristics Death Myocardial infarction Repeat revascularization Coronary artery bypass graft surgery Percutaneous coronary intervention Repeat hospitalization Revascularization procedures (per 100 patients) Coronary artery bypass graft surgery Percutaneous coronary intervention Diagnostic catheterizations (per 100 patients) Hospital admissions (per 100 patients) Hospital days (per 100 patients) Follow-up costs ($) Repeat procedures Hospital room/ancillary Inpatient physician fees Outpatient services/physician fees Total follow-up costs Aggregate 1-yr costs ($)
76 ⫾ 35 220 ⫾ 112 1.3 ⫾ 0.6 1.4 ⫾ 0.8 1.5 ⫾ 0.7 1.4 ⫾ 0.8 1,222 ⫾ 313 (1,190) 1,026 ⫾ 582 (939) 4,865 ⫾ 1,936 (3,700) 131 ⫾ 35 (126) 7,251 ⫾ 2,142 (6,643) 0.2% 2.3% 0.2% 0.2% 0.4% 1.7% 1.7 ⫾ 1.4 (1) 1.2 ⫾ 0.9 (1)
TAXUS-IV14 BMS (n ⫽ 522)
77 ⫾ 39 204 ⫾ 105 1.3 ⫾ 0.7 1.3 ⫾ 0.7 1.5 ⫾ 0.8 1.4 ⫾ 0.6 1,237 ⫾ 529 (1,163) 975 ⫾ 557 (929) 2,052 ⫾ 738 (1,700) 131 ⫾ 59 (123) 4,395 ⫾ 1,263 (4,302) 0.0% 1.5% 0.0% 0.0% 0.2% 2.3% 1.7 ⫾ 1.2 (1) 1.2 ⫾ 0.6 (1)
Paclitaxel-Eluting Stents (n ⫽ 662) 51 ⫾ 27 — — — 2.9 ⫾ 2.4 1.3 ⫾ 0.7
BMS (n ⫽ 652) 51 ⫾ 28 — — — 2.8 ⫾ 2.4 1.3 ⫾ 0.8
— — 3,966 ⫾ 1,363 (3,380)
— — 1,924 ⫾ 1,193 (1,440)
6,324 ⫾ 2,188 (5,740)
4,336 ⫾ 2,427 (3,852)
0.0% 2.4% 0.3% 0.0% 1.2% 1.6% 2.0 ⫾ 2.0 (1) 1.3 ⫾ 0.9 (1)
0.0% 2.1% 0.2% 0.2% 0.3% 1.9% 1.9 ⫾ 2.1 (1) 1.3 ⫾ 1.6 (1)
7,251 ⫾ 2,142 (6,643) 2,701 ⫾ 1,997 (1,832) 1,379 ⫾ 367 (1,284) 11,345 ⫾ 3,211 (10,462)
4,395 ⫾ 1,263 (4,302) 2,704 ⫾ 1,865 (1,832) 1,362 ⫾ 193 (1,284) 8,464 ⫾ 2,497 (7,884)
6,324 ⫾ 2,188 (5,740) 2,882 ⫾ 1,858 (2,497) 1,889 ⫾ 340 (1,749) 11,096 ⫾ 3,195 (10,165)
4,336 ⫾ 2,427 (3,852) 2,849 ⫾ 1,960 (2,497) 1,883 ⫾ 586 (1,749) 9,067 ⫾ 3,387 (8,230)
1.1% 0.8% 13.3% 1.3% 12.4% 27.0% 16.3 1.3 15.0 12.8 40.3 109.0
0.8% 1.9% 28.4% 3.0% 26.9% 41.1% 35.4 3.0 32.4 18.5 64.8 148.2
2.1% 1.1% 6.6% 1.7% 5.1% 18.6% 6.9 1.7 5.3 — 26.4 71.3
1.4% 2.5% 16.6% 3.8% 13.3% 26.4% 19.2 3.8 15.3 — 38.0 104.9
— — — — 3,487 (540) 14,583 (11,699)
— — — — 4,944 (90) 14,011 (9,540)
789 1,936 1,280 1,463 5,468 16,813
(0) (0) (0) (444) (2,522) (14,000)
1,788 3,149 1,759 1,343 8,040 16,504
(0) (0) (0) (728) (4,125) (12,802)
Data are expressed as mean ⫾ SD, numbers, or percentages. Values in parentheses are medians. Adapted with permission from Circulation13 and J Am Coll Cardiol.14 * Personnel costs include only nonphysician personnel. † In the SIRIUS trial, 1,058 patients were randomized. However, only 1,055 patients were included in the corresponding cost-effectiveness study.
procedure and adverse events, resource use, and costs of the index hospitalization and accrued up to 1 year, including the cost per repeat revascularization, are listed in Table 2. As in the SIRIUS trial, randomization to DES was associated with higher in-hospital costs during index hospitalization, whereas
BMS were associated with higher 1-year follow-up costs, resulting in similar aggregate costs at 1 year. To identify the incidence of LST (including very late stent thrombosis) in patients using DES, we conducted a systematic search of PubMed (National Library of Medi-
Coronary Artery Disease/Economic Impact of Late Stent Thrombosis Table 3 The effect of the cost of adverse events related to late stent thrombosis on the cost-effectiveness of drug-eluting stents Variable Incidence of LST (%) LST death (%) LST nonfatal myocardial infarction (%) Cost per death ($) Cost per nonfatal myocardial infarction ($) Cost of 2 additional years of clopidogrel ($) Cost per revascularization avoided ($) SIRIUS TAXUS-IV Cost per QALY gained ($) SIRIUS TAXUS-IV
Best-Case Scenario
Likely Scenario
0.2 40 60
0.5 40 60
Worst-Case Scenario 0.7 40 60
0 15,000
20,000 20,000
50,000 25,000
2,429
2,429
2,429
14,618 24,540
15,056 25,210
15,830 26,396
243,638 250,739
250,935 257,591
263,840 269,708
cine, Bethesda, Maryland) using the key words “DES” and “stent thrombosis.” We restricted our search to articles in English presenting pooled analyses comparing the incidence of LST in patients with DES to that in patients with BMS. We also restricted our search to studies that included ⱖ1,500 patients. Our search identified 5 studies, which reported increases in the incidence of LST that ranged from 0.2% to 0.7%.3,4,25,26,31 Therefore, we estimated the increase in the incidence of LST to be the middle of the range (0.5%) in patients receiving DES compared with patients receiving BMS, and we varied this value to the extremes of the reported range in sensitivity analyses. We also estimated that 40% of all episodes of LST were fatal, with the remaining 60% consisting of nonfatal myocardial infarctions. This case fatality rate is consistent with that observed in the SIRIUS trial. We calculated the ICER, expressed in cost per QALY gained, and the cost per revascularization avoided from the perspective of a third-party payer by incorporating the costs associated with LST into the cost-effectiveness analyses of the SIRIUS and TAXUS-IV trials. The costs associated with LST included the cost of nonfatal myocardial infarction, which we estimated to be $20,000 per nonfatal myocardial infarction, which is the approximate cost of repeat percutaneous coronary intervention,27,28 and the costs of death. To estimate the cost of death, we assumed that 50% of patients died suddenly with no treatment costs and that 50% of patients died after the initiation of treatment. We assumed that patients who died after the initiation of treatment had substantially higher treatment costs than those with nonfatal myocardial infarctions. Consequently, we assigned an average cost of $20,000 per death. In light of the new treatment guidelines recommending extended clopidogrel therapy in patients with DES,19 we also incorporated 2 additional years of clopidogrel therapy at a cost of $2,428 per DES patient.29 These costs were added to the cost per QALY gained and cost per revascularization originally presented in the SIRIUS and TAXUS-IV trials to obtain the
341
trial-specific relative increase in cost per patient. We then used these relative increases to obtain an ICER and cost per revascularization avoided that included the costs associated with LST. In sensitivity analyses, we examined the bestand worst-case scenarios by varying the increased LST incidence from 0.2% to 0.7%, the cost per death from $0 to $50,000, and the cost per myocardial infarction from $15,000 to $25,000. All costs are presented in United States dollars. Results Using an incidence of LST of 0.5% and a case fatality rate of 40%,26 2 additional deaths and 3 additional myocardial infarctions per 1,000 patients treated with DES are expected compared with the number of events originally reported in DES trials. Assuming a corresponding cost of $20,000 per death and $20,000 per myocardial infarction, and incorporating the cost of extended clopidogrel therapy, these additional events are associated with an increased cost of $2,510,600 per 1,000 patients treated with DES. The inclusion of these LST-associated costs results in a relative increase in cost per patient of 912% in the SIRIUS trial and 539% in the TAXUS-IV trial (Table 3). The costs associated with LST have important implications for measures of the cost-effectiveness of DES (Table 3). When these costs are taken into consideration, the cost per revascularization avoided increased from $1,650 in the SIRIUS trial to $15,056 and from $4,678 in the TAXUS-IV trial to $25,210. Similarly, LST-associated costs increased the cost per QALY gained from $27,500 in the SIRIUS trial to $250,935 and from $47,798 in the TAXUS-IV trial to $257,591. To examine the robustness of our estimates, we conducted sensitivity analyses in which we varied key parameters to produce best-case and worst-case scenarios (Table 3). This included varying the (1) the incidence of LST from 0.2% to 0.7%, (2) the cost per death from $0 to $50,000, and (3) the cost per myocardial infarction from $15,000 to $25,000. The effects of varying the incidence of LST from 0.2% to 0.7% are depicted in Figures 1 and 2. Under the assumptions for a best-case scenario, the cost per revascularization avoided was $14,618 in the SIRIUS trial and $24,540 in the TAXUS-IV trial. The cost per QALY gained in the best-case scenario was $243,638 in the SIRIUS trial and $250,739 in the TAXUS-IV trial. Under the assumptions for a worst-case scenario, the cost per revascularization avoided was $15,830 in the SIRIUS trial and $26,396 in the TAXUS-IV trial. The cost per QALY gained in the worst-case scenario was $263,840 in the SIRIUS trial and $269,708 in the TAXUS-IV trial. Discussion Our study was designed to examine the effect of LST on the cost-effectiveness of DES. Using data from the SIRIUS and TAXUS-IV trials, we found that LST, although rare, has a substantial effect on the cost-effectiveness of DES. The costs associated with LST increased the cost per revascularization avoided to $15,056 using data from the SIRIUS trial and to $25,210 using data from the TAXUS-IV trial.
342
The American Journal of Cardiology (www.AJConline.org)
Figure 1. Effect of varying the incidence of LST on the ICER in cost per QALY gained. The incidence of LST was varied from 0.2% to 0.7%, representing the range identified in a review of the published research. These sensitivity analyses are based on a cost of $20,000 per death, $20,000 per myocardial infarction, and 2 additional years of clopidogrel therapy in patients with DES. The dashed lines represent the accepted threshold for cost-effectiveness of $50,000 per QALY and the threshold for the gray zone of cost-effectiveness of $100,000 per QALY.
Similarly, LST-associated costs increased the ICER to $250,935 and $257,591 per QALY gained, respectively. Consequently, when the LST-associated costs are incorporated, DES exceeds the accepted thresholds of $100,000 per QALY (the limit of the gray zone of cost-effectiveness) and $10,000 per revascularization avoided, and thus, DES is no longer cost-effective. This finding was present using data from the SIRIUS and TAXUS-IV trials and for the 2 measures of cost-effectiveness. The costs associated with LST and corresponding increases in the ICER and cost per revascularization avoided have important policy implications. Previous studies have suggested that DES, although not cost effective from a hospital perspective, are cost effective from a societal perspective.21 However, the costs associated with LST shift measures of cost-effectiveness beyond the accepted thresholds. With limited resources available for health care and the advent of expensive treatment options (e.g., internal cardiac defibrillators, left ventricular assist devices) that are marginally cost effective, these findings confirm that the widespread use of DES is not justifiable.20 These findings are particularly important for the United States, where until recently, DES were used in ⬎80% of percutaneous coronary intervention procedures.30 Importantly, with no difference in mortality, previous cost-effectiveness analyses were based almost entirely on the observed difference in quality of life. Consequently, these analyses involved trading cost for improved quality of life in the absence of LST. However, once LST is incorpo-
rated, cost-effectiveness analyses now involve trading cost for improved quality of life and small but important increases in death and nonfatal myocardial infarction. The controversy surrounding DES has led to many costeffectiveness analyses.6 –18 In the United States, only 2 prospective cost-effectiveness analyses have been conducted.13,14 In these 2 studies, the investigators concluded that DES are cost effective in high-risk patients, on the basis of the cost per revascularization avoided. Cohen et al13 also reported an ICER in SIRIUS that is considered to be cost effective in the United States. The cost-effectiveness of DES has also been examined in 5 studies from Canada,6,7,15–17 as well as studies from Australia,8 Sweden,9 Switzerland,10 and the United Kingdom.11,18 All 5 Canadian studies found that DES were far from cost effective when used in a widespread, indiscriminate manner and may be only borderline cost effective in a very select percutaneous coronary intervention patient subgroup at the highest risk for restenosis. Cost-effectiveness analyses from Europe and Australia provided results that were similar to those reported in Canadian studies. These studies include the BAsel Stent Kosten Effektivitäts Trial (BASKET), which attempted to reproduce a “real-world” scenario by not implementing protocol-mandated follow-up angiography.10 The BASKET investigators concluded that in a real-world setting, DES should be reserved for patients in high-risk groups. The investigators also raised concerns regarding thrombotic complications associated with DES in the 6 months after the cessation of clopidogrel. In light of
Coronary Artery Disease/Economic Impact of Late Stent Thrombosis
343
Figure 2. Effect of varying the incidence of LST on the cost per revascularization avoided. The incidence of LST was varied from 0.2% to 0.7%, representing the range identified in a review of the published research. These sensitivity analyses are based on a cost of $20,000 per death, $20,000 per MI, and 2 additional years of clopidogrel therapy in patients with DES. The dashed line represents the accepted threshold for cost-effectiveness of $10,000 per revascularization avoided.
their concerns regarding thrombotic complications, the investigators suggested a need for prolonged clopidogrel therapy. Importantly, as with the other cost-effectiveness analyses of DES, the BASKET trial did not consider the costs of LST or the associated need for extended clopidogrel therapy. These additional costs render DES not cost effective in the United States, the only setting to date where the widespread use of DES may have been cost effective. The inclusion of these costs in cost-effectiveness analyses from other countries would further highlight the unfavorable cost-effectiveness of the widespread use of DES. Our cost-effectiveness analysis has a number of potential limitations. First, we only examined costs from a third-party payer perspective and did not included the societal costs associated with the loss of productivity or the humanistic cost associated with the additional loss of life. As a result, we underestimated the costs associated with LST, and thus, our ICER and cost per revascularization avoided represent conservative estimates. Second, we used conservative assumptions (e.g., incidence of 0.5%, 40% mortality) and conservative LST-associated costs (e.g., $20,000 per myocardial infarction, $20,000 per death). We also assumed that all repeat revascularizations were percutaneous coronary interventions rather than bypass surgeries, which would increase the cost of treating LST-associated events. It is therefore likely that we underestimated the costs associated with LST. Third, we also only included the cost of an additional 2 years of clopidogrel therapy. The inclusion of the cost associated with indefinite dual-antiplatelet therapy
would further increase the ICER and cost per revascularization avoided of DES. Fourth, it is possible that sustained clopidogrel therapy would reduce the observed rate of LST. However, in the extreme scenario of an LST incidence rate of 0%, the cost of clopidogrel alone renders DES not cost effective. Our analysis is based on aggregate published data from the SIRIUS and TAXUS-IV trials. Consequently, we were unable to examine the cost-effectiveness of DES among various subgroups of patients. Thus, although the widespread use of DES is no longer cost effective, their costeffectiveness in high-risk patients remains unknown. In addition, as is true with most trials, the generalizability of their results is unclear. Our analysis is also limited by the limitations of the original SIRIUS and TAXUS-IV costeffectiveness analyses. These limitations include minimizing the number of stents used per procedure and the use of protocol-mandated angiography. Minimizing the number of stents used per percutaneous coronary intervention results in a lower cost per repeat revascularization avoided and consequently a more favorable cost-effectiveness ratio. The inclusion of protocol-mandated angiography results in the reporting of higher rates of restenosis and repeat revascularization procedures in BMS patients and ultimately a more favorable estimate of the clinical effectiveness, and subsequently cost-effectiveness, of DES. These limitations of the SIRIUS and TAXUS-IV cost-effectiveness studies likely underestimate the true ICER and cost per revascularization
344
The American Journal of Cardiology (www.AJConline.org)
avoided of DES, exaggerating the cost-effectiveness of DES. 1. Bavry AA, Kumbhani DJ, Helton TJ, Borek PP, Mood GR, Bhatt DL. Late thrombosis of drug-eluting stents: a meta-analysis of randomized clinical trials. Am J Med 2006;119:1056 –1061. 2. Iakovou I, Schmidt T, Bonizzoni E, Ge L, Sangiorgi GM, Stankovic G, Airoldi F, Chieffo A, Montorfano M, Carlino M, et al. Incidence, predictors, and outcome of thrombosis after successful implantation of drug-eluting stents. JAMA 2005;293:2126 –2130. 3. Lagerqvist B, James SK, Stenestrand U, Lindback J, Nilsson T, Wallentin L. Long-term outcomes with drug-eluting stents versus baremetal stents in Sweden. N Engl J Med 2007;356:1009 –1019. 4. Stone GW, Moses JW, Ellis SG, Schofer J, Dawkins KD, Morice MC, Colombo A, Schampaert E, Grube E, Kirtane AJ, et al. Safety and efficacy of sirolimus- and paclitaxel-eluting coronary stents. N Engl J Med 2007;356:998 –1008. 5. Babapulle MN, Joseph L, Belisle P, Brophy JM, Eisenberg MJ. A hierarchical Bayesian meta-analysis of randomised clinical trials of drug-eluting stents. Lancet 2004;364:583–591. 6. Shrive FM, Manns BJ, Galbraith PD, Knudtson ML, Ghali WA. Economic evaluation of sirolimus-eluting stents. CMAJ 2005;172: 345–351. 7. Rinfret S, Cohen DJ, Tahami Monfared AA, Lelorier J, Mireault J, Schampaert E. Cost effectiveness of the sirolimus-eluting stent in high-risk patients in Canada: an analysis from the C-SIRIUS trial. Am J Cardiovasc Drugs 2006;6:159 –168. 8. Lord SJ, Howard K, Allen F, Marinovich L, Burgess DC, King R, Atherton JJ. A systematic review and economic analysis of drug-eluting coronary stents available in Australia. Med J Aust 2005;183:464 – 471. 9. Ekman M, Sjogren I, James S. Cost-effectiveness of the Taxus paclitaxel-eluting stent in the Swedish healthcare system. Scand Cardiovasc J 2006;40:17–24. 10. Kaiser C, Brunner-La Rocca HP, Buser PT, Bonetti PO, Osswald S, Linka A, Bernheim A, Zutter A, Zellweger M, Grize L, Pfisterer ME. Incremental cost-effectiveness of drug-eluting stents compared with a third-generation bare-metal stent in a real-world setting: randomised Basel Stent Kosten Effektivitats Trial (BASKET). Lancet 2005;366: 921–929. 11. Bagust A, Grayson AD, Palmer ND, Perry RA, Walley T. Cost effectiveness of drug eluting coronary artery stenting in a UK setting: cost-utility study. Heart 2006;92:68 –74. 12. Greenberg D, Bakhai A, Cohen DJ. Can we afford to eliminate restenosis? Can we afford not to? J Am Coll Cardiol 2004;43:513–518. 13. Cohen DJ, Bakhai A, Shi C, Githiora L, Lavelle T, Berezin RH, Leon MB, Moses JW, Carrozza JP Jr, Zidar JP, Kuntz RE. Cost-effectiveness of sirolimus-eluting stents for treatment of complex coronary stenoses: results from the Sirolimus-Eluting Balloon Expandable Stent in the Treatment of Patients With De Novo Native Coronary Artery Lesions (SIRIUS) trial. Circulation 2004;110:508 –514. 14. Bakhai A, Stone GW, Mahoney E, Lavelle TA, Shi C, Berezin RH, Lahue BJ, Clark MA, Lacey MJ, Russell ME, et al. Cost effectiveness of paclitaxel-eluting stents for patients undergoing percutaneous coronary revascularization: results from the TAXUS-IV trial. J Am Coll Cardiol 2006;48:253–261. 15. Bowen J, Hopkins R, He Y, Blackhouse G, Lazzam C, Tu JV, Cohen E, Tarride JE, Goeree R. Systematic Review and Cost-Effectiveness
16.
17.
18.
19.
20. 21. 22.
23. 24.
25.
26.
27.
28. 29.
30. 31.
Analysis of Drug Eluting Stents Compared to Bare Metal Stents for Percutaneous Coronary Interventions in Ontario: Interim Report for the Ontario Ministry of Health and Long-Term Care. Report No. HTA002– 0512. Hamilton, Ontario, Canada: Ontario Ministry of Health and Long-Term Care, 2005. Mittmann N, Brown A, Seung SJ, Coyle D, Cohen E, Brophy J, Title L, Oh P. Economic Evaluation of Drug Eluting Stents. Report No. 53 Ottawa, Ontario, Canada: Canadian Coordinating Office for Health Technology Assessment, 2005. Brophy J, Erickson L. An Economic Analysis of Drug Eluting Coronary Stents: A Quebec Perspective. Report No. 04-04. Montreal, Quebec, Canada: Agence d’Evaluation des Technologies et des Modes d’Intervention en Sante, 2004. Liverpool Reviews & Implementation Group. Coronary Artery Stents Rapid Systematic Review and Economic Evaluation (Addendum B). Commissioned by NHS R&D HTA Programme on Behalf of the National Institute for Clinical Excellence. Liverpool, United Kingdom: Liverpool Reviews & Implementation Group, 2003. King SB III, Smith SC Jr, Hirshfeld JW Jr, Jacobs AK, Morrison DA, Williams DO, Feldman TE, Kern MJ, O’Neill WW, Schaff HV, et al. 2007 focused update of the ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2008;51:172–209. Eisenberg MJ. Drug-eluting stents: the price is not right. Circulation 2006;114:1745–1754. Ryan J, Cohen DJ. Are drug-eluting stents cost-effective? It depends on whom you ask. Circulation 2006;114:1736 –1743. Gaziano TA, Opie LH, Weinstein MC. Cardiovascular disease prevention with a multidrug regimen in the developing world: a cost-effectiveness analysis. Lancet 2006;368:679 – 686. Weinstein MC, Stason WB. Foundations of cost-effectiveness analysis for health and medical practices. N Engl J Med 1977;296:716 –721. Cohen DJ, Taira DA, Berezin R, Cox DA, Morice MC, Stone GW, Grines CL. Cost-effectiveness of coronary stenting in acute myocardial infarction: results from the Stent Primary Angioplasty in Myocardial Infarction (Stent-PAMI) trial. Circulation 2001;104:3039 –3045. Kastrati A, Mehilli J, Pache J, Kaiser C, Valgimigli M, Kelbaek H, Menichelli M, Sabate M, Suttorp MJ, Baumgart D, et al. Analysis of 14 trials comparing sirolimus-eluting stents with bare-metal stents. N Engl J Med 2007;356:1030 –1039. Mauri L, Hsieh WH, Massaro JM, Ho KK, D’Agostino R, Cutlip DE. Stent thrombosis in randomized clinical trials of drug-eluting stents. N Engl J Med 2007;356:1020 –1029. Yock CA, Boothroyd DB, Owens DK, Garber AM, Hlatky MA. Cost-effectiveness of bypass surgery versus stenting in patients with multivessel coronary artery disease. Am J Med 2003;115:382–389. Nagle PC, Smith AW. Review of recent US cost estimates of revascularization. Am J Manag Care 2004;10(suppl):S370 –S376. Cowper PA, Udayakumar K, Sketch MH Jr, Peterson ED. Economic effects of prolonged clopidogrel therapy after percutaneous coronary intervention. J Am Coll Cardiol 2005;45:369 –376. Maisel WH. Unanswered questions— drug-eluting stents and the risk of late thrombosis. N Engl J Med 2007;356:981–984. Spaulding C, Daemen J, Boersma E, Cutlip DE, Serruys PW. A pooled analysis of data comparing sirolimus-eluting stents with bare-metal stents. N Engl J Med 2007;356:989 –997.