Cost-Effectiveness of Prasugrel Versus Clopidogrel in Patients With Acute Coronary Syndromes and Planned Percutaneous Coronary Intervention
Results From the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel–Thrombolysis in Myocardial Infarction TRITON-TIMI 38
Background— In patients with acute coronary syndromes and planned percutaneous coronary intervention, the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel–Thrombolysis in Myocardial Infarction 38 (TRITON-TIMI 38) demonstrated that treatment with prasugrel versus clopidogrel was associated with reduced rates of cardiovascular death, MI, or stroke and an increased risk of major bleeding. We evaluated the cost-effectiveness of prasugrel versus clopidogrel from the perspective of the US healthcare system by using data from TRITON-TIMI 38.
Methods and Results— Detailed resource use data were prospectively collected for all patients recruited from 8 countries (United States, Australia, Canada, Germany, Italy, Spain, United Kingdom, and France; n=3373 prasugrel, n=3332 clopidogrel). Hospitalization costs were estimated on the basis of diagnosis-related group and in-hospital complications. Cardiovascular medication costs were estimated by using net wholesale prices (clopidogrel=$4.62/d; prasugrel=$5.45/d). Life expectancy was estimated from in-trial cardiovascular and bleeding events with the use of statistical models of long-term survival from a similar population from the Saskatchewan Health Database. Over a median follow-up of 14.7 months, average total costs (including study drug) were $221 per patient lower with prasugrel (95% confidence interval, −759 to 299), largely because of a lower rate of rehospitalization involving percutaneous coronary intervention. Prasugrel was associated with life expectancy gains of 0.102 years (95% confidence interval, 0.030 to 0.180), primarily because of the decreased rate of nonfatal MI. Thus, compared with clopidogrel, prasugrel was an economically dominant treatment strategy. If a hypothetical generic cost for clopidogrel of $1/d is used, the incremental net cost with prasugrel was $996 per patient, yielding an incremental cost-effectiveness ratio of $9727 per life-year gained.
Conclusion— Among acute coronary syndrome patients with planned percutaneous coronary intervention, treatment with prasugrel versus clopidogrel for up to 15 months is an economically attractive treatment strategy.
Clinical Trial Registration— clinicaltrials.gov. Unique identifier: NCT00097591.
Received August 10, 2009; accepted October 29, 2009.
The goal of pharmacological management of patients with acute coronary syndromes (ACS) is to limit atherothrombotic events while minimizing bleeding. Dual antiplatelet therapy with aspirin and clopidogrel is the current mainstay for the treatment of ACS, including those undergoing percutaneous coronary intervention (PCI). The Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel–Thrombolysis in Myocardial Infarction 38 (TRITON-TIMI 38) demonstrated that treatment with prasugrel, a new thienopyridine, versus clopidogrel was associated with a significantly reduced rate of the composite end point of cardiovascular death, nonfatal myocardial infarction (MI), or nonfatal stroke at the expense of an increased risk of TIMI major bleeding not related to coronary artery bypass graft surgery (CABG).1
Clinical Perspective on p 79
The extent to which prasugrel should be considered the thienopyridine of choice for ACS patients undergoing PCI depends not only on the balance between thrombotic and bleeding events but also on the economic implications, especially in light of the potentially large population of patients who may be candidates for this new therapy. We therefore compared the overall costs of treatment with prasugrel versus clopidogrel and evaluated the cost-effectiveness of these therapies from the perspective of the US healthcare system in terms of incremental cost per year of life gained.
The design and key results of the TRITON-TIMI 38 trial have been reported previously.1,2 Briefly, TRITON-TIMI 38 was an international multicenter randomized controlled trial among 13 608 patients with moderate- to high-risk ACS undergoing planned PCI that compared prasugrel (60-mg loading dose and 10-mg daily maintenance dose) with clopidogrel (300-mg loading dose and a 75-mg daily maintenance dose) for 6 to 15 months. The primary end point was the composite of death resulting from cardiovascular causes, nonfatal MI, or nonfatal stroke. The key safety end point was TIMI major bleeding not related to CABG.
Design of the Economic Study
The economic analysis was prespecified and performed using data from all patients enrolled in 8 prespecified countries: United States, Australia, Canada, Germany, Italy, Spain, United Kingdom, and France (the necessary data to characterize resource use were collected only for patients from these countries). The primary economic end point was total in-trial costs; if one drug proved to be both more effective and more costly than the other, an evaluation of incremental cost-effectiveness would be performed.
Sources of Cost Data
The general approach to estimating costs was to multiply counts of resource use (hospitalizations, physician services, procedures, medications) by price weights derived from comparable populations of US patients. All costs other than study drug costs were assessed in 2005 US dollars.
Detailed rehospitalization data (indication for hospitalization, length of stay, diagnostic/therapeutic procedures performed) were collected for all end-point–related hospitalizations for patients enrolled in each of the 8 countries participating in the economic study. Data relating to hospitalizations for other cardiovascular or bleeding events were obtained from adverse event reporting. The index hospitalization and all rehospitalizations for a cardiovascular or bleeding condition were assigned a diagnosis-related group (DRG) on the basis of principal diagnosis and procedure(s) performed by a coder blinded to treatment assignment; costs were assigned on the basis of the average DRG-specific Medicare reimbursement rates. Physician costs associated with hospitalizations were estimated from the DRG-specific ratio of physician reimbursement to hospital costs.3–5
Because DRG-based costs are insensitive to differences in resource use that do not result in changes in DRG assignment, separate incremental costs were assigned to reflect the occurrence of procedure-related MI and bleeding complications. These costs were based on published regression-based cost analyses from similar US PCI patients enrolled in the Randomized Evaluation of PCI Linking Angiomax to Reduced Clinical Events (REPLACE-2) trial (periprocedural MI, $2543; TIMI major or minor bleed, $7176; TIMI minimal bleed, $451).6 Similarly, the incremental cost of a CABG-related major bleed ($9366) was derived from an analysis of Medicare data.7
Study drug costs were assigned through the use of net wholesale price as of August 2009 (clopidogrel, $4.62/d for the base case; prasugrel, $5.45/d).8 A secondary analysis comparing prasugrel with generic clopidogrel (anticipated to be available in 2011 in the United States) estimated the cost of generic clopidogrel to be $1/d. Because outpatient concomitant medication use was virtually identical between the 2 treatment groups, associated costs were not included.
Life Expectancy Estimation
The median duration of follow-up for patients enrolled in the economic study was 14.7 months for both groups. Because this duration was relatively brief compared with overall life expectancy for the trial population, estimation of life expectancy for each surviving TRITON patient beyond his or her observed follow-up duration, contingent on the clinical events (cardiovascular death, MI, stroke) observed during the trial, was required. These estimates were derived from an analysis of the Saskatchewan Health Database,9 a publicly available, comprehensive, longitudinal healthcare use database containing the entire population of the Canadian province of Saskatchewan. The Saskatchewan life expectancy estimates are based on data from patients who underwent PCI either during or after a hospitalization for ACS that occurred between 1985 and 1995. Follow-up for these patients ranged from <1 to 15 years (average, 3.35 years).
The estimation of life expectancy after a nonfatal MI or nonfatal stroke was based on data from Saskatchewan patients who experienced these nonfatal events within 1 year of the index ACS hospitalization to mirror the trial time frame. Methods used to derive the life expectancy estimates have been previously published10 and validated,11 and this approach has been used for the life expectancy projections for several previous trial-based economic analyses.4,5,12 Parametric regression was used to estimate piecewise hazard functions of death over time for patients who experienced nonfatal MI or stroke, with adjustment for age, sex, cardiovascular risk factors, and prior events/procedures. Life expectancy for each study participant who survived to the end of the study was predicted contingent on in-trial cardiovascular events. Life-years lost as a result of nonfatal MI or nonfatal stroke were estimated by subtracting the covariate-specific life expectancy estimates for patients with these events from the life expectancy of otherwise identical patients who experienced neither event within 1 year after PCI. Life-years lost as a result of death were estimated as the covariate-specific life expectancy of those who survived the year.
Categorical data are reported as frequencies; continuous data are reported as mean±SD. The clinical end points for patients in the economic study and for patients in the overall trial were compared between treatment groups through the use of unadjusted Cox proportional-hazards models. Rates of periprocedural MI and bleeding events during the index hospitalization were compared between treatment groups with the Pearson χ2 test; the Mantel-Haenszel test for trend was used to compare the incidence of multiple PCI and MI events over the course of follow-up. Poisson regression was used to compare rates of follow-up hospitalizations. Confidence intervals (CIs) for the differences in costs and projected life expectancy between the 2 treatment groups were obtained with the bootstrap method.13 All analyses were performed according to intention to treat.
Cost-effectiveness was evaluated from the perspective of the US healthcare system by using a lifetime horizon; the primary end point was the incremental cost per life-year gained. The incremental cost-effectiveness ratio (ICER) for prasugrel versus clopidogrel was calculated by dividing the net cost associated with prasugrel treatment (over a median duration of 14.7 months) by the difference in lost life expectancy resulting from death, nonfatal MI, or nonfatal stroke. Estimates of future annual healthcare costs beyond the trial period ($6836/y) were derived from the Reduction of Atherothrombosis for Continued Health Registry.14 Future costs and life expectancy were discounted at 3% annually in the base case.
Sensitivity analyses examined the impact of alternative assumptions about costs, life expectancy, prognostic impact of in-trial events, and discount rate (0%, 5%) on the results. The base-case analysis was repeated for several clinically relevant subgroups, some prespecified (sex, diabetes, ST-elevation versus non–ST-elevation ACS, patients receiving drug-eluting stents versus no stents) and some identified post hoc (history of stroke/transient ischemic attack [TIA], age ≥75 years, body weight <60 kg). To explore the impact of duration of therapy on cost-effectiveness, separate analyses were carried out with the use of cost and clinical outcomes data from randomization through 30 days and from 31 days through the end of follow-up.
We performed a secondary cost–utility analysis in which health benefits were assessed in terms of quality-adjusted life-years. Age- and sex-specific utility values derived from the Beaver Dam Health Outcomes Study15 were applied to all patients enrolled in the economic study, and the overall average was used as the baseline utility for all patients. A multiplicative utility adjustment of 0.88 for the first year of follow-up was applied to patients who experienced an in-trial MI, and a multiplicative utility adjustment of 0.52 for the patient’s remaining lifetime was made for patients who experienced an in-trial stroke.16,17 A disutility “toll” of 0.015 years was subtracted from the patient’s total quality-adjusted life expectancy for each major bleeding episode.18
For each analysis, bootstrap methods (5000 replicates) were used to estimate the proportion of the joint distribution of lifetime costs and effectiveness differences lying in different regions of the cost-effectiveness plane. Bootstrap methods also were used to produce cost-effectiveness acceptability curves.19
Role of the Study Sponsor
The analytical plan was prespecified in the economic study protocol and was performed by the authors without input from the study sponsor through the use of an independent copy of the complete trial database. The funding agreement stipulated that the investigators had the right to publish the study findings regardless of the results.
Patient Population and Clinical Results
Baseline characteristics were similar for the overall trial population and the economic study subset (see Table I of the online-only Data Supplement), as were the primary clinical results (Figure 1 and Table II of the online-only Data Supplement). There was a slightly greater relative hazard reduction with prasugrel early in follow-up (over the first 30 days) for the overall trial population (hazard ratio, 0.77; 95% CI, 0.67 to 0.88) than for the economic study subset (hazard ratio, 0.79; 95% CI, 0.66 to 0.96), whereas from 30 days through the end of the study, the estimated impact of prasugrel was greater for the economic study subset (hazard ratio, 0.81 [95% CI, 0.65 to 1.01] versus 0.87 [95% CI, 0.74 to 1.02]).
Index Hospitalization Resource Use and Costs
The distribution of DRGs associated with the index hospitalization was similar between treatment groups. The incidence of periprocedural MI events occurring during the index hospitalization was higher for clopidogrel (5.3% versus 4.6%; P=0.19), whereas the incidence of bleeding events was greater for prasugrel (5.3% versus 4.5%; P=0.13). Increased costs of $38 per patient related to periprocedural bleeding events for patients in the prasugrel arm were partially offset by decreased costs associated with periprocedural MIs of $19. As a result, mean index hospitalization costs did not differ significantly between treatment groups (mean difference, −$12; 95% CI, −174 to 156; Figure 2, left).
Follow-Up Resource Use and Costs
The incidence of rehospitalization for cardiovascular or bleeding indications was 36.4 per 100 patients with prasugrel and 37.7 per 100 patients with clopidogrel (P=0.39). Repeat PCI represented 43% of all rehospitalization events and was less frequent with prasugrel (14.1 versus 17.7 per 100 patients; P=0.0002; Table 1). Although the majority of these procedures were not associated with a new MI, the relative reduction with prasugrel was greater for PCI with MI than for PCI without MI (38% versus 16%). The absolute reduction, however, was greater for PCI without MI (12.0 versus 14.4 events per 100 patients) compared with PCI with MI (2.1 versus 3.4 events per 100 patients). Prasugrel also was associated with a significant reduction in the incidence of multiple PCI or MI events (68 [2.0%] versus 81 [2.4%] and 7 [0.21%] versus 17 [0.51%] patients with >1 and >2 PCI- or MI-related hospitalizations, respectively; P=0.0002 for test of trend). The rate of hospitalization for bleeding was higher for prasugrel (3.4 versus 2.3 per 100 patients; P=0.0079); the majority of such hospitalizations were due to gastrointestinal hemorrhage.
Total rehospitalization costs, exclusive of study drug, were $517 per patient lower for prasugrel (95% CI, 25 to 1040; Table 2); the majority of the cost offsets were due to the reduced rate of rehospitalization involving PCI (Figure 2, right). On the other hand, mean costs of study drug were $308 higher for prasugrel (95% CI, 269 to 347). Cumulative medical care costs (including study drug and the initial and follow-up hospitalizations) were thus $221 per patient lower with prasugrel (95% CI, −759 to 299).
Life Expectancy Projections
Lost life expectancy resulting from cardiovascular death, nonfatal MI, or nonfatal stroke was lower for prasugrel (0.428 versus 0.530 life-years lost; difference between prasugrel and clopidogrel (ΔP-C)=−0.102; 95% CI, −0.1803 to −0.0299), primarily because of the reduction in nonfatal MI (Table 3). When these projections were based on data for the overall trial cohort, life expectancy gains with prasugrel were slightly lower (0.096 life-years gained; 95% CI, 0.0426 to 0.1482).
Under our base-case assumptions, treatment with prasugrel versus clopidogrel for a median of 14.7 months was projected to decrease cost by $221 and to increase life expectancy by 0.102 years, rendering treatment with prasugrel an economically dominant strategy. Figure 3 illustrates the joint distribution of incremental cost and life expectancy differences derived from bootstrap resampling. Treatment with prasugrel was a dominant strategy in 79.7% of bootstrap replicates, and the ICER was <$50 000 per life-year gained in 99.8%. In our base-case cost–utility analysis, the overall benefit of prasugrel was 0.095 quality-adjusted life-year, and prasugrel remained a dominant strategy.
Impact of Duration of Therapy on Cost-Effectiveness
When we limited our analysis to costs and events that occurred during the initial 30 days of follow-up, prasugrel was associated with cost offsets of $218 (excluding study drug), net cost savings of $192 per patient (after including study drug), and a projected 0.056-year gain in life expectancy. In a landmark analysis beginning on day 31 of follow-up (ie, costs and events that occurred before this time point were excluded), prasugrel treatment was associated with cost offsets of $311 (excluding study drug), net cost savings of $29 per patient (after including study drug), and a projected 0.053-year gain in life expectancy. Thus, under our base-case assumptions, prasugrel remained an economically dominant strategy for both the initial 30 days of treatment and the extension of treatment from 31 days to 14.7 months.
Prasugrel remained a dominant therapy over a broad range of assumptions, including the application of hospitalization costs derived from the Marketscan database (private-payer perspective rather than Medicare); the inclusion of life-years lost resulting from all deaths (as opposed to only cardiovascular deaths and fatal bleeds); the inclusion of life-years lost due to TIMI major bleeding events (under the assumption that the associated lost life expectancy is equal to that for a nonfatal MI); the reduction in lost life expectancy resulting from death, MI, or stroke to 50% of that estimated from the Saskatchewan Health Database; the assumption of no life expectancy lost as a result of periprocedural MIs; and the application of alternative discount rates (see Table III of the online-only Data Supplement). The inclusion of cardiovascular healthcare costs in added years of life resulted in an incremental lifetime cost for prasugrel of $283 per patient and an associated ICER of $2768 per life-year gained (Table III of the online-only Data Supplement). In this scenario, the ICER remained <$10 000 per life-year gained in 83.0% of bootstrap replicates and <$50 000 per life-year gained in 99.5% of replicates.
Impact of Generic Clopidogrel
When prasugrel was compared with generic clopidogrel at a hypothetical cost of $1/d, the net cost with prasugrel increased to $996 per patient with an associated ICER of $9727 per life-year gained, and 98.2% of bootstrap estimates were <$50 000 per life-year gained (Figure 4). When this analysis was restricted to the initial 30 days of treatment, prasugrel was an economically dominant strategy with net cost savings of $92 per patient. Treatment from day 31 to the end of the trial yielded an estimated incremental cost with prasugrel of $1088 and a corresponding ICER of $20 714 per life-year gained with 84.5% of bootstrap estimates <$50 000 per life-year gained (Figure 4). Under the most conservative set of assumptions (inclusion of lost life expectancy as a result of TIMI major bleeding, use of generic price for clopidogrel, and inclusion of costs in added years of life), the ICER for prasugrel was $18 458 with >91% of bootstrap estimates <$50 000 per life-year gained.
Figure 5 illustrates the impact of a continuous range of differences in drug price between prasugrel and clopidogrel on the cost-effectiveness results. For treatment over the full study duration, the ICER remained <$30 000 per life-year gained even when the difference in drug price was as high as $10/d. For treatment up to 30 days, prasugrel remained a dominant strategy as long as the difference in drug price was <$7.67/d, whereas for treatment beyond 30 days, the ICER exceeded $50 000 per life-year gained only when the drug price difference was >$9.30/d.
Under our base-case assumptions, prasugrel was a dominant treatment strategy in an analysis of both costs and effectiveness based on US patients only. Prasugrel was also a dominant strategy in multiple other subgroups, including men, patients with or without diabetes, patients with ST-elevation or non–ST-elevation ACS, patients treated with bare metal stents, and patients with no history of stroke/TIA (Table 4). For women, there was a net increase in costs of $376 per patient and an ICER of $4001 per life-year gained. For patients treated with drug-eluting stents, prasugrel increased overall costs by $26, with a corresponding ICER of $326 per life-year gained. For patients with a history of stroke or TIA, the higher rate of primary clinical events with prasugrel translated into greater life expectancy losses.
This is the first prospectively designed economic evaluation of prasugrel compared with clopidogrel for the treatment of patients with ACS. On the basis of the 6705 ACS patients with planned PCI included in the TRITON-TIMI 38 health economic study, treatment with prasugrel versus clopidogrel for a median of 14.7 months appears to be an economically dominant treatment strategy, resulting in both lower costs and greater life expectancy. Prasugrel remained dominant in numerous sensitivity analyses varying either unit costs applied to the in-trial period or assumptions about the projected impact of in-trial events on life expectancy and quality of life.
The primary clinical benefit from prasugrel, with respect to the individual components of the primary clinical end point, was the reduction in nonfatal MI, which yielded 73% of the projected gain in life expectancy. The primary economic benefit from prasugrel was the reduced rate of rehospitalization for repeat PCI, which yielded >$600 per patient in cost offsets. Although the majority (83%) of these repeat procedures were not associated with a new MI, the relative reduction with prasugrel was greater for those associated with an MI. Incremental costs associated with the increase in bleeding events with prasugrel were more modest ($107 per patient) as a result of the relative infrequency of those events.
There has been considerable discussion in the literature about the relationship between bleeding and long-term survival among patients with coronary heart disease. Although numerous studies have demonstrated increased risk of death for patients with bleeding events,20–22 it remains uncertain whether this is due to a truly causal association or to common risk factors (ie, unmeasured confounding). Although major bleeding indicators were not available in the Saskatchewan Health Database from which to estimate the impact of bleeding events on life expectancy, a sensitivity analysis assuming that TIMI major bleeds have the same impact on life expectancy as nonfatal MIs showed that prasugrel remained a dominant strategy, although the associated life expectancy gain decreased from 0.102 years (base-case analysis) to 0.081 years.
We also performed a series of landmark analyses examining cost-effectiveness separately for the first 30 days of the trial and beyond 30 days. The 30-day landmark was chosen because it was considered a practical time point at which continuation of therapy with prasugrel after PCI might be reconsidered in the interest of maintaining the balance between thrombosis prevention and bleeding. Cost-effectiveness results showed prasugrel to be a dominant strategy both during the initial 30 days and from 31 days throughout the rest of the follow-up period. Although the validity of the results for the period beyond 30 days is uncertain because of the lack of rerandomization of patients at 30 days, it seems reasonable to expect that such an analysis would likely underestimate the true benefit of prasugrel because of the absence of a possible rebound effect early in the post–30-day period for patients who switch from prasugrel to clopidogrel (rather than stay on clopidogrel) at 30 days. Furthermore, the higher rate of multiple rehospitalizations for MI or repeat PCI for clopidogrel raises the possibility that prasugrel reduces the rate of subsequent hospitalization events beyond the first.23
Although prasugrel was a dominant therapy within 13 of the 17 subgroups examined, it is noteworthy that cost savings and estimated incremental life expectancy gains were considerably greater for the highest-risk subgroups (eg, patients with diabetes and patients with ST-elevation myocardial infarction). On the other hand, post hoc subgroup analyses from TRITON-TIMI 38 revealed that for patients with a prior history of stroke or TIA, there was a trend toward worse clinical outcomes in terms of both the primary clinical end point and bleeding and that these events translated into lower life expectancy for prasugrel (by 0.14 years). Therefore, despite lower net costs of $154 per patient, prasugrel would not be considered a therapy of choice for these patients.
Generic clopidogrel is likely to become available in the United States in 2011. Consideration of the impact of a reduced price for clopidogrel as low as $1.00/d yielded a net in-trial cost of treatment with prasugrel of $977 per patient and an ICER of $9727 per life-year gained. Under this hypothetical scenario, prasugrel remained an economically dominant strategy when treatment was restricted to the first 30 days of follow-up. Although treatment beyond 30 days was no longer economically dominant, the projected ICER of $20 714 per life-year gained compares favorably with many other accepted medical interventions.
Although this cost-effectiveness study derives strength from being carried out using patient-level data from the TRITON-TIMI 38 trial, it is not without limitations. TRITON-TIMI-38 was a multinational study, and healthcare cost data were not collected directly from the patients. Moreover, details sufficient to allow accurate accounting of subsequent hospitalizations were collected for only the 8 countries participating in the economic substudy. The effects of prasugrel versus clopidogrel on the relative incidence of the primary clinical end point and its individual components were similar between the overall trial cohort and the economic study subset; however, we were unable to rigorously assess whether the rates of non–end point–related hospitalizations differed. The lower rate of hospitalizations involving PCI in the absence of an MI accounts for a large portion of the cost offsets with prasugrel, and whether this difference between treatment groups is present among patients not in the economic study is unknown. The application of US DRG-based cost estimates to multinational hospitalizations does not fully account for possible differences in treatment practices and resource use between countries or healthcare systems. If a large proportion of patients come from countries for which the threshold for hospitalization differs considerably from that in the United States, the difference in costs between treatment arms may be underestimated or overestimated. Estimates of the incremental costs associated with periprocedural bleeds and MIs were derived from external sources; the definitions associated with these events may differ from those used in TRITON-TIMI 38.
The validity of the external estimates of lost life expectancy from the Saskatchewan Health Database depends on the nonfatal events on which they are based having the same prognostic implication as those observed within each of the treatment arms of the trial, an assumption that cannot be independently tested with available data. However, sensitivity analyses demonstrated that the cost-effectiveness results remained favorable under plausible variations in life expectancy projections, including the assumption that lost life expectancy resulting from death, nonfatal MI, or nonfatal stroke was only 50% of that estimated using Saskatchewan data and under the assumption that periprocedural MIs do not affect long-term prognosis.
For patients undergoing PCI in the ACS setting, treatment with prasugrel compared with clopidogrel is highly cost-effective, and under many circumstances cost saving, over both the subacute and longer-term phases of treatment. Further studies are necessary to understand the optimal duration of treatment and the cost-effectiveness in other non-ACS settings.
We thank Pat McCollam, PharmD, of Eli Lilly & Co, Inc, for his support throughout all phases of this study.
Source of Funding
This health economic study was funded by a research grant to Drs Mahoney and Cohen from Eli Lilly & Co, Inc.
Dr Mahoney has received grant support from Eli Lilly, Daiichi Sankyo, Sanofi-Aventis, and Bristol-Myers Squibb and has received lecture fees from Sanofi-Aventis and Bristol-Myers Squibb. Dr Cohen has received grant support from Sanofi-Aventis, Bristol-Myers Squibb, Eli Lilly, Daiichi Sankyo, Schering-Plough, The Medicines Co, Accumetrics, Cordis, and Boston Scientific. Drs Wiviott, Antman, and Braunwald have received research grants from Daiichi Sankyo, Eli Lilly, and Sanofi-Aventis. In addition, Dr Braunwald has received consulting or paid advisory board fees from Daiichi Sankyo and Sanofi-Aventis and lecture fees from Eli Lilly and Sanofi-Aventis. Dr Wiviott has received consulting or advisory board fees from Daiichi Sankyo and Sanofi-Aventis and lecture fees from Eli Lilly and Daiichi Sankyo. Dr Antman has received consulting or advisory board fees from Sanofi-Aventis and lecture fees from Eli Lilly and Sanofi-Aventis. The other authors report no conflicts.
Wiviott SD, Antman EM, Gibson CM, Montalescot G, Riesmeyer J, Weerakkody G, Winters KJ, Warmke JW, McCabe CH, Braunwald E. Evaluation of prasugrel compared with clopidogrel in patients with acute coronary syndromes: design and rationale for the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel Thrombolysis in Myocardial Infarction 38 (TRITON-TIMI 38). Am Heart J. 2006; 152: 627–635.
Mitchell J, Burge R, Lee A, McCall NT. Per Case Prospective Payment for Episodes of Hospital Care. Springfield, Va: US Department of Commerce National Technical Information Service; 1995.
Mahoney EM, Mehta S, Yuan Y, Jackson J, Chen R, Gabriel S, Lamy A, Culler S, Caro J, Yusuf S, Weintraub WS. Long-term cost-effectiveness of early and sustained clopidogrel therapy for up to 1 year in patients undergoing percutaneous coronary intervention after presenting with acute coronary syndromes without ST-segment elevation. Am Heart J. 2006; 151: 219–227.
Cohen DJ, Lincoff AM, Lavelle TA, Chen HL, Bakhai A, Berezin RH, Jackman D, Sarembock IJ, Topol EJ. Economic evaluation of bivalirudin with provisional glycoprotein IIB/IIIA inhibition versus heparin with routine glycoprotein IIB/IIIA inhibition for percutaneous coronary intervention: results from the REPLACE-2 trial. J Am Coll Cardiol. 2004; 44: 1792–1800.
Downey W, Beck P, McNutt M, Stang MR, Osei W, Nichol J. Health databases in Saskatchewan. In: Strom BL, ed. Pharmacoepidemiology. 3rd ed. West Sussex, UK: John Wiley and Sons Ltd; 2002: 325–345.
Caro JJ, Migliaccio-Walle K, We JJ. Can we predict survival following a myocardial infarction sufficiently well. Eur Heart J. 2007; 28 (suppl 1): S89. Abstract.
Chen J, Bhatt DL, Dunn ES, Shi C, Caro JJ, Mahoney EM, Gabriel S, Jackson JD, Topol EJ, Cohen DJ. Cost-effectiveness of clopidogrel plus aspirin versus aspirin alone for secondary prevention of cardiovascular events: results from the CHARISMA Trial. Value Health. 2009; 12: 872–879.
Mahoney EM, Wang K, Cohen DJ, Hirsch AT, Alberts MJ, Eagle K, Mosse F, Jackson JD, Steg G, Bhatt DL. One-year costs in patients with a history of or at risk for atherothrombosis in the United States. Circ Cardiovasc Qual Outcomes. 2008; 1: 38–45.
Fryback DG, Dasbach EJ, Klein R, Klein BE, Dorn N, Peterson K, Martin PA. The Beaver Dam Health Outcomes Study: initial catalog of health-state quality factors. Med Decis Making. 1993; 13: 89–102.
Tsevat J, Goldman L, Soukup JR, Lamas GA, Connors KF, Chapin CC, Lee TH. Stability of time-tradeoff utilities in survivors of myocardial infarction. Med Decis Making. 1993; 13: 161–165.
Naglie IG, Detsky AS. Treatment of chronic nonvalvular atrial fibrillation in the elderly: a decision analysis. Med Decis Making. 1992; 12: 239–249.
Eikelboom JW, Mehta SR, Anand SS, Xie C, Fox KA, Yusuf S. Adverse impact of bleeding on prognosis in patients with acute coronary syndromes. Circulation. 2006; 114: 774–782.
Manoukian SV, Feit F, Mehran R, Voeltz MD, Ebrahimi R, Hamon M, Dangas GD, Lincoff AM, White HD, Moses JW, King SB III, Ohman EM, Stone GW. Impact of major bleeding on 30-day mortality and clinical outcomes in patients with acute coronary syndromes: an analysis from the ACUITY Trial. J Am Coll Cardiol. 2007; 49: 1362–1368.
Murphy SA, Antman EM, Wiviott SD, Weerakkody G, Morocutti G, Huber K, Lopez-Sendon J, McCabe CH, Braunwald E. Reduction in recurrent cardiovascular events with prasugrel compared with clopidogrel in patients with acute coronary syndromes from the TRITON-TIMI 38 trial. Eur Heart J. 2008; 29: 2473–2479.
The Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel–Thrombolysis in Myocardial Infarction 38 (TRITON-TIMI 38) showed that for patients with acute coronary syndromes and planned percutaneous coronary intervention, prasugrel was associated with a significantly reduced rate of the composite end point of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke at the expense of an increased risk of TIMI major bleeding not related to coronary artery bypass graft surgery. The extent to which prasugrel should be considered the thienopyridine of choice for acute coronary syndrome patients undergoing percutaneous coronary intervention, however, depends not only on the balance between thrombotic and bleeding events but also on the economic implications, especially in light of the potentially large population of patients who may be candidates for this therapy. This study uses data from TRITON to evaluate the cost-effectiveness of prasugrel versus clopidogrel from the perspective of the US healthcare system. Using current drug prices of $4.62/d for clopidogrel and $5.45/d for prasugrel, we found that treatment with prasugrel versus clopidogrel for a median of 14.7 months was associated with lower costs and life expectancy gains both during the initial 30 days and from 31 days throughout the rest of the trial follow-up period, both overall and for most subgroups studied, including high-risk subgroups of patients with diabetes or ST-segment–elevation myocardial infarction. Results from the comparison of prasugrel with generic clopidogrel, likely to become available in the United States in 2011, at a hypothetical price for clopidogrel of $1/d yielded an overall increase in costs during the trial period of $996 per patient and an associated incremental cost-effectiveness ratio of $9727 per life-year gained, still favorable compared with many other accepted interventions.
Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.
Guest Editor for this article was Veronique L. Roger, MD, MPH.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.109.900704/DC1.