Drug-Eluting Stents

Abstract

HomeCirculationVol. 114, No. 16Drug-Eluting Stents Free AccessArticle CommentaryPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessArticle CommentaryPDF/EPUBDrug-Eluting StentsThe Price Is Not Right Mark J. Eisenberg, MD, MPH Mark J. EisenbergMark J. Eisenberg From the Divisions of Cardiology and Clinical Epidemiology, Jewish General Hospital/McGill University, Montreal, Quebec, Canada. Search for more papers by this author Originally published17 Oct 2006https://doi.org/10.1161/CIRCULATIONAHA.106.646190Circulation. 2006;114:1745–1754“Does it make economic sense to completely abandon a therapy that works well for 85% to 90% of the population for a new therapy costing four times as much to treat a transient health condition with no impact on either death or myocardial infarction?”J.M. Brophy and L.J. Erickson1Since Andreas Gruntzig performed the first percutaneous coronary intervention (PCI) in 1977,2 the cost-effectiveness of this procedure has engendered major controversy. Debates have erupted over the clinical value and cost-effectiveness of each new device or therapy that has become available. Controversies have arisen regarding the cost of atherectomy, bare metal stents (BMS), brachytherapy, distal protection devices, glycoprotein IIb/IIIa inhibitors, and intravascular ultrasound.3–6 Drug-eluting stents (DES) are the most recent devices to have their cost scrutinized.7–12Response by Ryan and Cohen p 1754Clinical Effectiveness and DES PenetrationBalloon angioplasty is associated with restenosis rates of 30% to 40%, whereas PCI with BMS is associated with rates of 20% to 30%,13,14 and PCI with DES is associated with rates in the single digits.15,16 My colleagues and I pooled the results of 11 DES trials involving >5000 patients using a hierarchical Bayesian random-effects model.17 We found that, compared with BMS, DES reduce angiographic restenosis from 29.3% to 8.9% (Table 1 and Figure 1). There was no difference between DES and BMS in terms of mortality (0.9% versus 0.9%, respectively) or myocardial infarction (2.7% versus 2.9%, respectively). There was a suggestion that restenosis was less with sirolimus-eluting stents (SES) compared with polymeric paclitaxel-eluting stents (PES) (6.2% for SES versus 36.9% for BMS; 7.1% for PES versus 23.5% for BMS), a finding that was subsequently identified in another meta-analysis.18TABLE 1. Clinical Events and Restenosis Rates in Randomized Clinical Trials Investigating DESsTrialAngiographic Restenosis, %Repeat Revascularization, %*Myocardial Infarction, %†Mortality, %MACE, %DESBMSDESBMSDESBMSDESBMSDESBMSMACE indicates major adverse cardiac events; RAVEL, Randomized study with the sirolimus-eluting Bx Velocity balloon-expandable stent (Cypher); ASPECT, ASian Paclitaxel-Eluting stent Clinical Trial; ELUTES, European evaLUation of pacliTaxel Eluting Stent; and PATENCY, PAclitaxel-eluting sTENt for CYtostatic prevention of restenosis.*Revascularization by repeat PCI or coronary artery bypass surgery of the index lesion, including for stent thrombosis.†Both Q wave and non–Q wave.‡Cardiac deaths only.§These are the correct rates; there was an error in the original article.¶The MACE rate in this study included only cardiac deaths.Adapted and reproduced from Babapulle et al17 with permission from Elsevier. Copyright 2004.Sirolimus RAVEL026.3022.93.34.21.71.75.829.7 SIRIUS8.936.34.116.62.83.20.90.67.118.9 C-SIRIUS2.351.14.018.02.04.0004.018.0 E-SIRIUS5.942.34.020.94.62.31.10.68.022.6 Pooled6.236.93.518.53.23.21.00.76.821.0Paclitaxel, polymeric TAXUS I010.3010.000003.210.0 TAXUS II7.121.94.214.43.15.30‡0.8‡10.421.7 TAXUS IV7.926.63.011.33.53.71.4‡1.1‡8.5§15.1§ Pooled7.123.53.312.23.34.00.91.08.716.7Paclitaxel, nonpolymeric ASPECT8.027.36.83.42.61.70.908.55.2 ELUTES13.121.67.215.81.300.709.918.4 DELIVER16.722.45.26.81.41.01.01.06.6¶8.6¶ PATENCY38.135.312.519.20003.812.523.1 Pooled14.823.86.07.61.50.90.90.97.79.5Sirolimus and paclitaxel pooled8.929.34.213.22.72.90.90.97.88.9Download figureDownload PowerPointFigure 1. Forest plot comparing rates of angiographic restenosis for DES and BMS. Reproduced from Babapulle et al17 with permission from Elsevier. Copyright 2004.The interventional community quickly embraced the results of the DES trials. DES use has become nearly ubiquitous in the United States,19 and its use is becoming widespread outside the United States as well. Rather than reserving this high-cost technology for patients who are at high risk for restenosis, many interventional cardiologists are placing these stents in all patients, including those whose baseline risk of restenosis is low. Before the universal use of DES becomes an entrenched practice, we need to know the answer to the following question: Is the clinical benefit associated with DES substantial enough to justify the use of this high-cost technology in all patients undergoing PCI? Several lines of evidence suggest that DES are currently too expensive to be used in an across-the-board manner in all patients undergoing PCI. These data come from a variety of studies comparing the cost-effectiveness of DES and BMS that have been performed in various countries.Cost-Effectiveness of DES Versus BMSA PubMed search for studies comparing the cost-effectiveness of DES versus BMS identified 7 studies from North America (Table 2)20–26 and 6 studies from Australia and Europe (Table 3).27–32 Only studies that reported cost per quality-adjusted life-year (QALY) gained or cost per repeat revascularization avoided were included. Cost per QALY gained is the primary outcome measure of most cost-effectiveness analyses. Using this measure, we can directly compare the cost-effectiveness of different healthcare interventions (Table 4).33 In the United States, an intervention associated with a cost per QALY gained of <$50 000 is considered to be cost-effective; one associated with a cost per QALY gained of $50 000 to $100 000 is in the “gray area”; and one associated with a cost per QALY gained of >$100 000 is considered unattractive.34 Cost per repeat revascularization avoided is the other measure that is commonly used in DES cost-effectiveness studies. This disease-specific measure allows us to examine whether the incremental cost of DES above that of BMS is offset by the cost savings brought about by a reduction in the need for subsequent revascularization procedures. In the United States, a cost per revascularization avoided <$10 000 is thought to be cost-effective.20,21TABLE 2. Cost-Effectiveness Studies of DES in North AmericaLocation and AuthorSource of Efficacy DataSource of Cost DataPatient Group, nAssumptionsOutcomes at 12 MonthsMean Stents, n (type)Revascularization. Rate, DES vs BMS, %Cost of DES vs Cost of BMSCost of PCI vs Cost of CABGCost per QALY GainedCost per Revascularization AvoidedCABG indicates coronary artery bypass grafting; QALY, quality-adjusted life-years; NR, not reported; CCN, Cardiac Care Network of Ontario; CARDIACCESS, patient registry of the CCN; DM, diabetes mellitus; MI, myocardial infarction; APPROACH, Alberta Provincial Project for Outcome Assessment in Coronary Heart Disease; RCTs, Randomized Controlled Trials; CHUM, Centre Hospitalier de l’Université de Montréal; and RAMQ, Régie de l’assurance maladie du Québec.*After follow-up of 2 years.†Complex includes all patients, including those with diabetes, long lesions, small vessels, and multivessel disease.‡Cost of BMS PCI (cost of SES PCI) including stent price.§After follow-up of 9 months.United States (costs in $US) Greenberg et al20BMS data, RAVEL, SIRIUSMulticenter trialsSingle vessel, 60001.3 (SES)3.8 vs 14.02700 vs 700NR vs 25 000NR7000* Cohen et al21SIRIUSHospital and Medicare ratesComplex†, 10581.4 (SES)16.3 vs 35.42900 vs 9004395 (7251)‡ vs NR27 5401650Canada (costs in $CAN) Bowen et al22CCN CARDIACCESSNRNo DM, no MI, 47961.5 (Both)9.4 vs 10.71899 vs 600≈7050 vs 18 7992 221 69295 383No DM, post-MI, 14321.5 (Both)10.0 vs 13.11899 vs 600≈7050 vs 18 7991 688 78669 696DM, no MI, 13771.5 (Both)11.1 vs 13.51899 vs 600≈7050 vs 18 7991 132 42649 333DM, post-MI, 3481.5 (Both)12.1 vs 16.91899 vs 600≈7050 vs 18 799438 41517 711 Shrive et al23APPROACHAPPROACH, Alberta HealthComplex†, 73341.4 (SES)NR2900 vs 50015 569 vs 32 00958 721NR Mittmann et al24Meta-analysisRCTs, Ontario Medicare, personal communicationComplex†, 24471.5 (PES)3.3 vs 12.22400 vs 6089761 vs 19 617NR26 562Complex†, 17481.5 (SES)3.5 vs 18.52400 vs 6089761 vs 19 617NR12 527 Brophy et al1,25Meta-analysis17Hospital costs, Quebec MedicareComplex†, 14001.7 (NR)3.3 vs 12.82600 vs 7004507 vs 15 025NR23 067§ Rinfret et al 26C-SIRIUSCHUM, RAMQSingle vessel, 1001.5 (SES)4.0 vs 22.02700 vs 7004006 vs 14 402NR11 275TABLE 3. Cost-Effectiveness Studies of DES in Australia and EuropeLocation and AuthorSource of Efficacy DataSource of Cost DataPatient Group, nAssumptionsOutcomes at 12 MonthsMean Stents, nRevascularization. Rate, DES vs BMS, %Cost of DES vs Cost of BMSCost of PCI vs Cost of CABGCost per QALY GainedCost per Revascularization AvoidedCABG indicates coronary artery bypass grafting; QALY, quality-adjusted life-years; NR, not reported; RAVEL, Randomized study with the sirolimus-eluting Bx Velocity balloon-expandable stent (Cypher); BASKET, Basel Stent Kosten Effektivitäts Trial; TARMED, Swiss Medical tariff; CTC, cardiothoracic center; NHS, National Health Service; and NICE, National Institute for Clinical Excellence.*Complex includes all patients, including those with diabetes, long lesions, small vessels, and multivessel disease.†High risk includes patients with diabetes, small vessels, and/or long lesions.‡Prices vary depending on the type of DES or BMS used.§After follow-up of 6 months.Australia (cost in $AUS) Lord et al27TAXUS, SIRIUS, RAVELNational Hospital Cost DataComplex*, 17481.5 (SES)4.0 vs 20.52000 vs 850NR46 8293746Complex*, 16421.5 (PES)4.5 vs 14.62000 vs 850NR76 4676117 Ward28Patient registryDepartment of HealthComplex*, 4901.1 (NR)0 vs 5Δ17005186 vs 18 496NR28 349Sweden (cost in SEK) Ekman et al29TAXUS IVHospital Price ListsComplex*, 13 2001.4 (PES)4.5 vs 15.9Δ960052 300-66 020 vs 134 5072 350 84446 801High risk†, NR1.4 (PES)6 vs 22Δ960052 300–66 020 vs 134 507381 5547648Switzerland (cost in ) Kaiser et al30BASKETTARMEDComplex*, 8261.4 (Both)7.2 vs 12.11935–2380‡ vs 1300-1260‡3095 vs 709573 283‡18 311§United Kingdom (cost in GBP) Bagust et al31CTC, RAVEL, SIRIUSCTC and NHS CostsElective PCI, 19511.9 (SES)7.5 vs 24.9Δ5003190 vs 7750NR51 600–238 900Nonelective PCI, 9331.7 (SES)NRΔ5004179 vs 7750NR−23 700–133 600 NICE32TAXUS II, RAVELNRSingle-vessel disease, NR1.0 (NR)2.7 vs 12.7900 vs 3802156 vs 836824 3251080TABLE 4. Cost-Effectiveness and Use of Selected Interventions in the Medicare Population*InterventionCost-Effectiveness†ImplementationPET indicates positron-emission tomography.*Ranges rather than point estimates are provided because the actual cost-effectiveness will vary according to the target populations and the strategies used.†Calculated as cost/QALY. The calculation was based on 2002 dollars.‡With the use of this intervention, benefits are lower and costs are higher than with the use of the standard workup.Adapted and reproduced from Neumann et al33 with permission from the Massachusetts Medical Society. Copyright 2005.Influenza vaccineCost saving40%–70%Pneumococcal vaccineCost saving55%–65%β-Blockers after myocardial infarction<10 000†85%Mammographic screening10 000–25 000†50%–70%Colon-cancer screening10 000–25 000†20%–40%Osteoporosis screening10 000–25 000†35%Management of antidepressant medicationCost saving up to 30 000†40%–55%Hypertension medication (diastolic blood pressure >105 mm Hg)10 000–60 000†35%Cholesterol management, as secondary prevention10 000–50 000†30%Implantable cardioverter-defibrillator30 000–85 000†100 000 cases/yDialysis in end-stage renal disease50 000–100 000†90%Lung-volume–reduction surgery100 000–300 000†10 000–20 000 cases/yLeft ventricular assist devices500 000–1.4 million†5000–100 000 cases/yPET in Alzheimer’s diseaseDominated‡50 000 cases/yCost-Effectiveness Studies of DES in the United StatesAt the time of this writing, only 2 traditional cost-effectiveness studies examining DES in the context of the United States have been published; both are from the same group at the Harvard Clinical Research Institute (Table 2).20,21 Despite these surprisingly limited cost-effectiveness data, DES are being used in the vast majority of PCI procedures now being performed in the United States.19Greenberg et al20 published a review of the economic impact of restenosis and DES. Embedded within the review was a decision-analytic model examining DES cost-effectiveness. The model used outcome and resource use data from >6000 “real-world” patients undergoing single-vessel PCI procedures.35–37 Costs were based on pooled data from several clinical trials involving >3000 patients. The model used the following assumptions: (1) a BMS repeat revascularization rate of 14%, (2) an 80% reduction in repeat revascularization rates with DES, (3) an incremental cost of $2000 per DES, and (4) a mean use of 1.3 stents per PCI. Over a 2-year follow-up, this model indicated that overall medical care costs are approximately $900 per patient higher with DES than with BMS, with an incremental cost-effectiveness ratio of approximately $7000 per repeat revascularization avoided. Sensitivity analyses suggested that treatment with DES is cost saving for patients with a BMS repeat revascularization rate >20% and cost-effective (<$10 000 per repeat revascularization avoided) for patients with a BMS repeat revascularization rate >12% (Figure 2). The authors concluded that, compared with BMS, DES are cost saving for only a modest proportion of the current PCI population in the United States. However, they did suggest that DES are economically attractive for virtually all diabetic patients and for nondiabetic patients with small vessels and long lesions. Greenberg et al did not report a cost per QALY gained, but they concluded that the cost-effectiveness of DES is dependent on the target population undergoing PCI and the alternative therapy that might be used (ie, medical therapy, BMS, or coronary artery bypass grafting. Download figureDownload PowerPointFigure 2. Relationship between the rate of target vessel revascularization with BMS implantation and the incremental cost-effectiveness of DES implantation for patients undergoing single-vessel PCI. Reproduced from Greenberg et al20 with permission from the American College of Cardiology Foundation. Copyright 2004.Cohen et al21 reported an economic analysis of the SIRIUS trial.2 Clinical outcomes, resource use, and costs were prospectively collected for 1058 patients who received either an SES or a BMS over a 1-year period. Initial hospital costs were increased by $2881 per patient with DES. During the 1-year follow-up, use of DES versus BMS was associated with reductions in the rates of repeat PCI (12.4% versus 26.9%, respectively) and bypass surgery (1.3% versus 3.0%, respectively). Although follow-up costs were reduced by $2571 per patient with DES, aggregate 1-year costs were still $309 per patient higher. The incremental cost-effectiveness ratios for DES were $27 540 per QALY gained and $1650 per repeat revascularization avoided. The authors concluded that, although the use of SES was not cost saving compared with BMS, for patients undergoing PCI of complex coronary lesions, the use of DES appeared to be reasonably cost-effective within the context of the United States healthcare system.These 2 studies examining the cost-effectiveness of DES in the context of the United States came to similar conclusions. Both studies suggested that DES are cost-effective in high-risk patients with respect to repeat revascularizations avoided. In addition, although the study by Greenberg et al20 did not report a cost per QALY gained, the figure reported by Cohen et al21 falls within the range that is generally accepted as being cost-effective in the United States (<$50 000 per QALY gained) (Table 4). However, the conclusions drawn from these 2 studies must be tempered in view of the assumptions and methodologies used.Greenberg et al20 assumed that all patients underwent single-vessel PCI with a mean use of 1.3 DES per procedure. In contrast, other cost-effectiveness studies assumed a mean use of 1.5 DES per PCI; some have even used 1.7 or 1.9 (Tables 2 and 3). Minimizing the number of DES used per PCI minimizes the estimated cost per repeat revascularization avoided. In addition, because DES allow us to treat more complex lesions than were treated previously, it is likely that future PCI procedures will use more rather than fewer stents per case.Cohen et al21 performed a textbook-perfect cost-effectiveness study as part of the SIRIUS trial. However, the results of this study were affected by the use of protocol-mandated angiography. Most of the DES trials, including the SIRIUS trial, used protocol-mandated follow-up angiography at 6 to 9 months with subsequent clinical follow-up several months later. Because follow-up angiography was performed in all patients, many cases of angiographic restenosis were identified in patients who were asymptomatic. If restenosis was identified at the time of the protocol-mandated angiography, repeat PCI was frequently performed—often called the oculostenotic reflex. These asymptomatic patients who had angiographic but not clinical restenosis were then identified as achieving one of the predefined end points of the trial: need for repeat revascularization. The effect of protocol-mandated angiography can be seen in the Kaplan-Meier survival curves for the RAndomized study with the sirolimus-eluting Bx VELocity balloon-expandable stent (Cypher; RAVEL) trial (Figure 3).38 At the time of the mandated angiographic follow-up, there was a sharp rise in the identification of restenosis and occurrence of repeat revascularization procedures. Repeat revascularization in the BMS group jumped from 6% before angiography to >20% after angiography. Similar increases were seen in other DES trials that used protocol-mandated angiography (Table 5).28,39–41 Although Cohen et al tried to account for the use of protocol-mandated angiography in their analysis, they were overly optimistic in both the clinical effectiveness data that they used in their cost analysis and the subsequent DES cost-effectiveness that they reported. Download figureDownload PowerPointFigure 3. Kaplan-Meier estimates of survival free of myocardial infarction and repeated revascularization among patients who received SES and those who received BMS in the RAVEL trial. The percent of event-free patients in the BMS group decreased rapidly during the period of angiographic follow-up. This decrease was due to an increase in restenosis and subsequent target lesion revascularizations identified by the angiographic follow-up. Adapted and reproduced from Morice et al with permission from the Massachusetts Medical Society. Copyright 2002.TABLE 5. Target Lesion Revascularization Rates With BMS in DES Trials Before and After Protocol-Mandated AngiographyTrialRevascularization Rate, %Before Protocol-Mandated AngiographyAfter Protocol-Mandated AngiographyAdapted and reproduced from Ward28 with permission from the Australasian Society of Cardiac and Thoracic Surgeons and the Cardiac Society of Australia and New Zealand. Copyright 2005.RAVEL623SIRIUS717E-SIRIUS1121TAXUS II515Cost-Effectiveness Studies of DES Outside the United StatesIn contrast to the 2 published studies from the United States, 5 studies from Canada suggest that across-the-board use of DES in all PCI patients is not cost-effective. Although the Canadian dollar is worth ≈90 cents American at the time of this writing, the thresholds considered to be cost-effective in Canada are somewhat different than those in the United States. The reason for this is that the costs of repeat PCI and coronary artery bypass grafting are much less in Canada.42 Cost-effectiveness thresholds in Canada are <Can $50 000 per QALY gained and <Can $12 551 per repeat revascularization avoided.26Bowen et al22 performed a cost-effectiveness analysis of DES for the Ontario Ministry of Health and Long-Term Care. These investigators found that DES were associated with an exceedingly high cost per QALY gained: Can $438 415 to Can $2 221 692, ratios that clearly place this technology in the non–cost-effective range. In addition, the cost per revascularization avoided was also prohibitive: from Can $17 711 for patients with a recent myocardial infarction and diabetes to Can $95 383 for patients without a myocardial infarction or diabetes.Shrive et al23 performed a cost-effectiveness analysis of SES on behalf of the Alberta Provincial Project for Outcome Assessment in Coronary Heart Disease (APPROACH) investigators. They found that SES use was associated with a cost per QALY gained of Can $58 721 and that SES use was more cost-effective in patients with diabetes and in those >75 years of age (Can $44 135 and Can $40 129 per QALY gained, respectively). For patients <65 years of age and those without diabetes, SES use was substantially less cost-effective (Can $72 464 and Can $63 383 per QALY gained, respectively).Mittmann et al24 performed a cost-effectiveness analysis of DES for the Canadian Coordinating Office for Health Technology Assessment. The investigators found that PES use was associated with a cost per revascularization avoided of Can $26 562 to Can $29 048, whereas SES use was associated with a cost of Can $12 527 to Can $16 600. The investigators did not calculate costs per QALY gained. However, they did examine the impact of DES use on the annual Canadian healthcare budget. They found that extending DES use from the 40% of patients at highest risk for restenosis to 100% of patients undergoing PCI would lead to a >3-fold increase in DES costs but only a 1.5-fold reduction in repeat revascularization procedures.Brophy and Erickson25 performed a cost-effectiveness analysis of DES for the Quebec Agency for the Evaluation of Technology and Health Interventions. The investigators calculated that cost per revascularization avoided would increase from Can $7000 at 20% DES penetration to Can $23 067 at 100% DES penetration. The investigators also calculated the price at which DES use would be cost neutral assuming different DES penetration rates. With a 20% use in patients at highest risk, the break-even cost for DES would be Can $1663; at 60%, it would be Can $1266; and at 100%, it would be Can $1161.Rinfret et al26 investigated the cost-effectiveness of SES versus BMS in high-risk patients with single long de novo lesions in small coronary arteries. These investigations found that BMS use versus balloon angioplasty is associated with a cost per repeat revascularization avoided of Can $12 551 and that SES versus BMS use was associated with a cost per repeat revascularization avoided of Can $11 275. The investigators concluded that DES are borderline cost-effective in Canada in a high-risk subgroup of patients.Thus, 5 cost-effectiveness studies from Canada suggest that DES are not an attractive therapy to be used in an across-the-board manner. The authors of each of these studies suggested that, at current prices, DES are too expensive to be cost-effective except in selected groups of high-risk patients.Authors of DES cost-effectiveness studies in Australia, Sweden, Switzerland, and the United Kingdom all reported results similar to those found in Canada (Table 3). Importantly, the BAsel Stent Kosten Effektivitäts Trial (BASKET) investigators prospectively performed a study in which they examined the cost-effectiveness of DES in a group of patients randomized to DES or BMS.30 As opposed to most other studies, costs were prospectively collected, and patients did not undergo protocol-mandated follow-up angiography. A total of 826 patients were randomized to SES, PES, or BMS. The aggregate costs at 6 months were higher with DES than with BMS, and higher stent costs were not compensated for by lower follow-up costs (overall 6-month costs were still 905 higher in the DES group). The incremental cost-effectiveness ratio to avoid 1 major adverse cardiac event was 18 311 , and the cost per QALY gained was more than 50 000 . The authors concluded that, in a real-world setting, use of DES should be restricted to patients in high-risk groups. Importantly, BASKET also suggested that DES are associated with a significantly higher rate of thrombotic complications compared with BMS during the 6 months following the cessation of clopidogrel.43 Therefore, the need for prolonged treatment with clopidogrel will further reduce the cost-effectiveness of DES.Limitations of DES Cost-Effectiveness StudiesDespite the fact that most studies have not found DES to be cost-effective in an across-the-board manner, many of these studies still painted an overly optimistic picture of the cost-effectiveness of DES. Two limitations of these studies were responsible. First, the true cost of DES procedures was underestimated; second, the clinical effectiveness of DES was overestimated.Although the cost of a single DES is reasonably well established, the published studies used a mean number of stents per PCI between 1.0 and 1.9. The assumption of a lower number of stents per PCI leads to a better cost-effectiveness ratio; the assumption of a higher number leads to a worse cost-effectiveness ratio. With the advent of DES, there has been a trend of performing increasingly complex PCI procedures. This trend will likely lead to an even greater number of stents per patient in the future. Thus, previous studies underestimated the true cost of DES procedures, and this led to an overly optimistic view of DES cost-effectiveness.Previous studies also overestimated the clinical effectiveness of DES. Most of these studies based their calculations on data obtained from previous DES trials. Unfortunately, most of these trials enrolled patients with solitary de novo coronary lesions and used protocol-mandated angiography. These features led to the observations of low restenosis rates in patients receiving DES and high rates of restenosis and repeat revascularization procedures in patients receiving BMS. The use of inflated estimates of clinical effectiveness led to overly optimistic estimates of DES cost-effectiveness. Studies like the BASKET trial that were performed in real-world settings without the use of protocol-mandated angiography found that the clinical benefit of DES is substantially less than that described in randomized controlled trials. Consequently, the cost-effectiveness studies that relied on the early DES trials were overly optimistic in the clinical effectiveness data that they used.Conclusions From DES Cost-Effectiveness StudiesDespite their limitations, several important conclusions can be derived from the DES cost-effectiveness studies that have been published. First, DES are not cost saving and they are not cost neutral. A substantial amount of money has to be spent to obtain a modest clinical benefit. Second, if DES are used in place of BMS, some but not all of the initial cost is recouped during the follow-up period as a result of a reduced need for repeat revascularization procedures. Third, because DES have no impact on mortality or myocardial infarction rates and because their effect on quality of life is modest, DES are associated with a high cost per QALY gained. Thus, the sole justification for the use of DES is their ability to reduce the need for repeat revascularization procedures. If these procedures are expensive, the incremental cost-effectiveness ratios associated with DES are attractive; if these procedures are inexpensive, the incremental cost-effectiveness ratios are unattractive. Finally, the cost-effectiveness of DES can be markedly enhanced by either decreasing their price or reserving their use for patients who are at high risk for restenosis. At current prices, using DES in an across-the-board manner is not an optimal strategy from a cost-effectiveness point of view.Other DES Cost StudiesBesides the traditional cost-effectiveness studies detailed above, a number of studies have examined the economics of DES from other perspectives. Several studies examined the decline in DES price required for DES to be cost neutral (the break-even price).44,45 Several studies examined the impact of DES penetration on hospital budgets,46,47 and other studies examined the impact of DES on global healthcare budgets.48–51 The conclusions from these studies can be summarized as follows. First, DES prices have to decline substantially before break-even prices are reached. Second, from both hospital and societal perspectives, across-the-board use of DES leads to substantial increases in budgetary costs that are not recouped by a reduced need for subsequent revascularization procedures.Competing Healthcare InterventionsAnother issue should be considered before deciding on a policy to guide our use of DES. Even in affluent societies, resources available for healthcare interventions are not unlimited.48 In reality, multiple potential healthcare interventions are in competition for the same resources; consequently, some interventions receive funding while others do not. The best way to determine which interventions will be funded is not by individual studies of cost-effectiveness but instead by directly contrasting the cost-effectiveness ratios of alternative interventions. Many healthcare interventions are associated with a cost per QALY gained in the gray area of $50 000 to $100 000. If we funded each of these interventions, we would quickly deplete the budget of