Although several drug-eluting stents (DESs) have been shown to be economically attractive compared to bare-metal stents in patients at moderate to high risk of restenosis, little is known about the cost–effectiveness of alternative DES designs, especially second-generation DESs. We therefore performed an economic substudy alongside the SPIRIT-IV trial, in which 3,687 patients undergoing single or multivessel percutaneous coronary intervention were randomized to receive second-generation everolimus-eluting stents (EESs; n = 2,458) or first-generation paclitaxel-eluting stents (PESs; n = 1,229). Costs through 2 years of follow-up were assessed from the perspective of the United States health care system. The primary cost-effectiveness end point was the incremental cost-effectiveness ratio assessed as cost per quality-adjusted life year gained. Over a 2-year period, use of EESs versus PESs led to a trend toward decreased overall repeat revascularization procedures (14.2 vs 16.2 per 100 subjects, p = 0.20) driven by a significant decrease in the number of target vessel revascularization procedures (8.2 vs 11.0 per 100 subjects, p = 0.02) but also a slight increase in the number of nontarget vessel revascularization procedures (6.0 vs 5.1 per 100 subjects, p = 0.37). Follow-up cardiovascular costs were decreased by $273/patient in the EES group (95% confidence interval for difference 1,048 less to 502 more, p = 0.49). Formal cost-effectiveness analysis based on these results demonstrated that the probability that EES was an economically attractive strategy (incremental cost-effectiveness ratio <$50,000/quality-adjusted life year gained) was 85.7%. These findings demonstrate that in patients undergoing percutaneous coronary intervention with DESs, use of EESs is economically attractive compared to PESs with improved clinical outcomes and lower overall medical care costs at 2 years.
The Clinical Evaluation of the XIENCE V Everolimus Eluting Coronary Stent System (SPIRIT)-IV trial—the largest head-to-head comparison of alternative drug-eluting stents (DESs)—demonstrated that compared to paclitaxel-eluting stents (PESs) everolimus-eluting stents (EESs) resulted in improved safety (lower risk of myocardial infarction and stent thrombosis) and lower rates of clinical restenosis. In contrast to most previous DES trials, SPIRIT-IV included patients undergoing multilesion and multivessel percutaneous coronary intervention without protocol-driven angiographic follow-up. By reflecting routine clinical practice more closely than most previous trials, the SPIRIT-IV trial provided an ideal opportunity to study the cost–effectiveness of alternate DES platforms. We therefore designed a prospective economic substudy alongside the SPIRIT-IV trial. The goals of the study were (1) to compare 2-year costs of percutaneous coronary intervention using EESs versus PESs; (2) to examine the cost–effectiveness of EESs versus PESs in patients undergoing percutaneous coronary intervention in a real-world setting (i.e., multiple lesions, no angiographic follow-up); and (3) to examine the impact of alternative device acquisition costs on the cost–effectiveness of EESs versus PESs.
Methods
The design and principal outcomes of the SPIRIT-IV trial ( NCT01016041 ) have been described previously. Briefly, this was a prospective, multicenter, randomized, single-blind controlled clinical trial in which 3,687 patients with angina or inducible ischemia and up to 3 previously untreated native coronary artery lesions (maximum 2 lesions/vessel) were randomly assigned, in a 2:1 ratio, to receive an EES (XIENCE V, Abbott Vascular, Santa Clara, California) or a PES (TAXUS Express , Boston Scientific, Natick, Massachusetts).
The analytic perspective of our study was that of the United States health care system, and all costs were assessed in 2010 United States dollars. Costs from years other than 2010 were inflated to 2010 dollars based on the medical care component of the Consumer Price Index. For this analysis, medical care costs from initial hospitalization through 2 years of follow-up were estimated using a combination of resource-based accounting (for procedure costs), regression modeling from a large single-center percutaneous coronary intervention database (for other costs during index hospitalization), and Medicare reimbursement rates (for subsequent cardiovascular hospitalizations and revascularization procedures) as previously described.
Because there were no differences between treatment groups in any measurements of resource use for the index percutaneous coronary intervention (i.e., procedure duration, contrast volume, balloons, guiding catheters, guidewires, etc.), these costs were excluded from our analysis. Stent costs were included, and acquisition costs of PES and EES were set at $1,933 per stent based on a national survey of United States hospitals (IMS Hospital Supply Index, accessed August 2009) for our base case analysis. Alternative stent acquisition costs were explored in sensitivity analyses.
All other costs for the initial percutaneous coronary intervention and hospitalization were calculated using a linear regression model based on all patients undergoing percutaneous coronary intervention at Saint Luke’s Mid America Heart Institute in 2009, with costs derived from the hospital’s microcost accounting system. In this multivariable model, total in-hospital cost (excluding cost of stents) was the dependent variable, and the following covariates were used as independent variables: age, gender, diabetes, previous coronary artery bypass grafting, previous percutaneous coronary intervention, acute coronary syndrome presentation, and in-hospital complications that included major bleeding (according to the American College of Cardiology National Cardiovascular Data Registry (NCDR) CathPCI Registry definition ), periprocedural myocardial infarction, stroke, and urgent repeat revascularization.
Follow-up hospitalization costs were estimated based on national reimbursement rates for similar clinical events. Each follow-up hospitalization was mapped to a diagnosis-related group using the Centers for Medicare and Medicaid Services grouper software, and costs were assigned based on mean Medicare reimbursement rates for fiscal year 2009. Costs of coronary angiography alone without revascularization were also included. Costs of follow-up hospitalizations that did not involve repeat revascularization or coronary angiography were excluded because these costs would be expected to be equal in the 2 arms.
Costs were assigned for physician services for inpatient procedures and daily care based on the Medicare fee schedule. Use of outpatient services including stress testing and physician office visits was not tracked during the study; therefore, these costs were excluded from the economic analysis. Medication use including duration of dual antiplatelet therapy was identical for the 2 groups over the 2-year follow-up period and thus excluded.
Categorical variables are reported as frequencies, and continuous data are reported as mean ± SD. Discrete variables were compared using chi-square or Fisher’s exact test. Normally distributed continuous variables were compared by Student’s t test. Cost and other non-normally distributed data (length of stay, procedure duration) were compared by Wilcoxon rank-sum test. All statistical analyses and cost–effectiveness analyses were performed according to the intention-to-treat principle.
The primary end point for the cost–effectiveness analysis was the incremental cost per quality-adjusted life year (QALY) gained. For this analysis, quality-adjusted life expectancy for each patient was estimated from a previous study that examined the association between repeat revascularization events and population-level utilities in 771 United States patients with percutaneous coronary intervention enrolled in the Stent Primary Angioplasty in Myocardial Infarction (Stent-PAMI) trial. We applied this disutility value to estimate the decrement in QALYs for all patients who had a repeat revascularization event. We assumed there would be no differences in long-term survival or costs beyond the 2-year observation period, and therefore any differences in quality-adjusted life expectancy were based entirely on observed differences in in-trial repeat revascularization events.
We also performed a secondary analysis whereby cost–effectiveness was assessed as incremental cost per repeat revascularization avoided. This “disease-specific” incremental cost–effectiveness ratio (ICER) was calculated by dividing the difference in 2-year mean medical care costs by the 2-year difference in the frequency of repeat revascularization related events (target or nontarget vessel revascularization) between the EES and PES groups. To estimate the uncertainty surrounding cost differences and cost–effectiveness ratios, we calculated bias-corrected confidence intervals (CIs) by the bootstrap method (1,000 replicates). Results of these analyses are displayed graphically as the joint distribution of the incremental cost and effectiveness of EES versus PES in the cost–effectiveness plane and as cost–effectiveness acceptability curves that indicate the probability that EES is cost–effective compared to PES over a range of alternative willingness-to-pay thresholds. For cost–effectiveness analyses, all future costs and outcomes were discounted by 3% per year.
We performed sensitivity analyses to explore the impact of alternative stent acquisition costs on our results. In addition, because we observed that the more effective stent (EES) was associated with an unexpected trend toward increased rates of nontarget vessel related revascularization procedures, we performed sensitivity analyses to more fully characterize the impact of these nontarget vessel revascularization events on the cost–effectiveness of EESs versus PESs.
Results
Baseline clinical and angiographic characteristics were well balanced between the 2 treatment groups ( Table 1 ). As previously described, the primary clinical outcome (target lesion failure: composite of cardiovascular death, target vessel myocardial infarction, and target lesion revascularization) was decreased by 30% at 2 years with EESs versus PESs (6.9% vs 9.9%, p = 0.003). These differences were driven by significant decreases in myocardial infarction (2.5% vs 3.9%, p = 0.02) and ischemia-driven target lesion revascularization (4.5% vs 6.9%, p = 0.004). In addition, stent thrombosis was less frequent with EESs (0.4% vs 1.2%, p = 0.008).
| Variable | EES | PES | p Value |
|---|---|---|---|
| (n = 2,458) | (n = 1,229) | ||
| Age (years) | 63 ± 10 | 63 ± 10 | 0.80 |
| Men | 67.8% | 67.8% | 1.00 |
| Race/ethnicity | 0.18 | ||
| White | 93.5% | 93.7% | |
| Black | 5.0% | 5.5% | |
| Other | 1.5% | 0.8% | |
| Diabetes mellitus | 32.0% | 32.5% | 0.92 |
| Current smoker | 21.4% | 21.9% | 0.62 |
| Previous myocardial infarction | 20.5% | 19.4% | 0.35 |
| Number of narrowed coronary arteries | 0.20 | ||
| 1 | 61.6% | 61.2% | |
| 2 | 27.9% | 29.9% | |
| 3 | 10.5% | 9.0% | |
| Number of arteries treated | 0.15 | ||
| 1 | 83.4% | 81.9% | |
| 2 | 15.8% | 17.7% | |
| 3 | 0.8% | 0.4% | |
| Number of narrowings treated | 0.54 | ||
| 1 | 75.2% | 74.7% | |
| 2 | 21.7% | 21.6% | |
| 3 | 3.1% | 3.7% |
Table 2 presents resource use and costs for the index revascularization procedure and subsequent hospital stay. In general, procedural resource use was comparable for the 2 groups. There were also no significant differences in hospital complications or resource use between the 2 treatment groups during the index percutaneous coronary intervention hospitalization. Initial hospital costs were similar for the EES and PES groups ($17,736 vs $17,758, p = 0.81).
| Variable | EES | PES | p Value |
|---|---|---|---|
| (n = 2,458) | (n = 1,229) | ||
| Index percutaneous coronary intervention procedure resource use | |||
| Procedure duration (min) | 42 ± 30 | 43 ± 32 | 0.83 |
| Number of balloons used | 1.6 ± 0.8 | 1.6 ± 0.8 | 0.13 |
| Number of study stents implanted | 1.5 ± 0.8 | 1.5 ± 0.8 | 0.32 |
| Bare-metal stents implanted | 1.1% | 1.1% | 0.82 |
| Glycoprotein IIb/IIIa used | 19.0% | 19.9% | 0.56 |
| Adverse events during index hospitalization | |||
| Death | 0% | 0% | — |
| Myocardial infarction | 1.5% | 1.8% | 0.45 |
| Repeat revascularization | 0.4% | 0.5% | 0.72 |
| Coronary artery bypass grafting | 0.1% | 0.0% | 0.56 |
| Percutaneous coronary intervention | 0.3% | 0.5% | 0.57 |
| Coronary angiography without revascularization | 0.3% | 0.4% | 0.77 |
| Stroke | 0.0% | 0.1% | 0.33 |
| Bleeding complication | 0.6% | 0.6% | 0.88 |
| Length of stay (days) | 1.4 ± 1.0 | 1.4 ± 0.9 | 0.71 |
| Index hospitalization cost ($) | 17,736 ± 2,636 | 17,758 ± 2,884 | 0.81 |
Resource use over the 2-year follow-up period and associated health care–related costs are presented in Table 3 . At 1-year follow-up, use of EESs compared to PESs led to a significant decrease in the number of target vessel-related revascularization procedures (4.6 vs 6.8 per 100 patients treated, 95% CI for difference 0.4 to 3.9, p = 0.01). By 2 years the target vessel revascularization benefit of EESs versus PESs had increased slightly with an absolute decrease in the total number of target vessel revascularization procedures of 2.9 events per 100 patients treated with EESs (95% CI 0.5 to 5.2, p = 0.02). There was no significant difference in the overall number of repeat revascularization procedures (14.2 vs 16.2, difference −2.0 procedures per 100 patients treated, 95% CI for difference −5.1 to 1.1, p = 0.20), however, owing to a small statistically insignificant increase in the number of nontarget vessel revascularization–related procedures (3.5 vs 2.7 events per 100 patients treated, mean difference 0.9 procedures, 95% CI for difference −1.0 to 2.7, p = 0.365). Projected quality-adjusted life expectancy was also slightly greater with EESs than with PESs (1.970 vs 1.964 QALYs, mean difference 0.006, p = 0.13).
| Cumulative Number of Events (per 100 patients) | 1-Year Results | 2-Year Results ⁎ | ||||||
|---|---|---|---|---|---|---|---|---|
| EES | PES | Difference (95% CI) † | p Value | EES | PES | Difference (95% CI) | p Value | |
| (n = 2,458) | (n = 1,229) | (n = 2,458) | (n = 1,229) | |||||
| Revascularization procedures | 8.1 | 9.4 | −1.3 (−3.5 to 0.9) | 0.23 | 14.2 | 16.2 | −2.0 (−5.1 to 1.1) | 0.20 |
| Coronary artery bypass grafting surgery | 1.1 | 0.98 | 0.1 (−0.6 to 0.8) | 0.73 | 1.8 | 1.9 | −0.1 (−1.0 to 0.9) | 0.93 |
| Percutaneous coronary intervention | 7.1 | 8.5 | −1.4 (−3.4 to 0.7) | 0.19 | 12.5 | 14.4 | −1.9 (−4.8 to 0.9) | 0.19 |
| Target vessel revascularization | 4.6 | 6.8 | −2.2 (−3.9 to −0.4) | 0.01 | 8.2 | 11.1 | −2.9 (−5.2 to −0.5) | 0.02 |
| Coronary artery bypass grafting surgery | 1.0 | 0.8 | 0.2 (−0.5 to 0.9) | 0.55 | 1.5 | 1.7 | −0.2 (−1.1 to 0.6) | 0.58 |
| Percutaneous coronary intervention | 3.6 | 5.9 | −2.3 (−3.9 to −0.8) | 0.004 | 6.8 | 9.4 | −2.5 (−4.7 to −0.4) | 0.02 |
| Nontarget vessel revascularization | 3.5 | 2.7 | 0.8 (−0.5 to 2.1) | 0.227 | 6.0 | 5.1 | 0.9 (−1.0 to 2.7) | 0.365 |
| Coronary angiography | 13.1 | 11.4 | 1.7 (−0.9 to 4.3) | 0.21 | 20.7 | 18.2 | 2.4 (−0.9 to 5.8) | 0.16 |
| Target vessel revascularization related | 919 | 1,328 | −409 (−786 to −32) | 0.03 | 1,627 | 2,288 | −661 (−1,181 to −142) | 0.013 |
| Nontarget vessel revascularization related | 1,620 | 1,305 | 315 (−33 to 663) | 0.08 | 2,699 | 2,289 | 411 (−79 to 901) | 0.10 |
| Total | 2,331 | 2,443 | −112 (−640 to 416) | 0.68 | 4,326 | 4,576 | −250 (−1,002 to 501) | 0.51 |
| All revascularization procedures included | 20,274 | 20,391 | −117 (−688 to 454) | 0.69 | 22,061 | 22,335 | −273 (−1,048 to 502) | 0.49 |
| Excluding nontarget vessel revascularization costs | 18,654 | 19,086 | −432 (−859 to −5) | 0.047 | 19,362 | 20,046 | −684 (−1,241 to −128) | 0.016 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
