Author + information
- Received August 1, 2011
- Revision received September 20, 2011
- Accepted October 11, 2011
- Published online January 10, 2012.
- Tracy Y. Wang, MD, MHS, MSc⁎,⁎ (, )
- Frederick A. Masoudi, MD, MSPH†,
- John C. Messenger, MD†,
- Kendrick A. Shunk, MD, PhD‡,
- Andrew Boyle, PhD, MBBS‡,
- J. Matthew Brennan, MD, MPH⁎,
- H. Vernon Anderson, MD§,
- Kevin J. Anstrom, PhD⁎,
- David Dai, PhD⁎,
- Eric D. Peterson, MD, MPH⁎,
- Pamela S. Douglas, MD⁎ and
- John S. Rumsfeld, MD, PhD∥
- ↵⁎Reprint requests and correspondence:
Dr. Tracy Y. Wang, Duke Clinical Research Institute, 2400 Pratt Street, Durham, North Carolina 27705
Objectives This study assessed the comparative effectiveness of drug-eluting stents (DES) versus bare-metal stents (BMS) among patients ≥85 years of age.
Background Despite an aging population, little is known about the comparative effectiveness of DES versus BMS among patients age ≥85 years undergoing percutaneous coronary intervention (PCI).
Methods We examined 471,006 PCI patients age ≥65 years at 947 hospitals in the National Cardiovascular Data Registry between 2004 and 2008 and linked to Medicare claims data. Long-term outcomes (median follow-up 640.8 ± 423.5 days) were compared between users of DES and BMS.
Results Patients age ≥85 years comprise an increasing proportion of PCIs performed among elderly subjects, yet rates of DES use declined the most in this age group. Compared with BMS, use of DES was associated with lower mortality: age ≥85 years, 29% versus 38% (adjusted hazard ratio [HR]: 0.80 [95% confidence interval (CI): 0.77 to 0.83]); age 75 to 84 years, 17% versus 25% (HR: 0.77 [95% CI: 0.75 to 0.79]); and age 65 to 74 years, 10% versus 16% (HR: 0.73 [95% CI: 0.71 to 0.75]). However, the adjusted mortality difference narrowed with increasing age (pinteraction <0.001). In contrast, the adjusted HR for myocardial infarction rehospitalization associated with DES use was significantly lower with increasing age: age ≥85 years, 9% versus 12% (HR: 0.77 [95% CI: 0.71 to 0.83]); age 75 to 84 years, 7% versus 9% (HR: 0.81 [95% CI: 0.77 to 0.84]); and age 65 to 74 years, 7% versus 8% (HR: 0.84 [95% CI: 0.80 to 0.88]) (pinteraction <0.001).
Conclusions In this national study of older patients undergoing PCI, declines in DES use were most pronounced among those aged ≥85 years, yet lower adverse-event rates associated with DES versus BMS use were observed.
With a progressively aging U.S. population (1), there is a need to evaluate the treatment of coronary heart disease and associated outcomes in older adults. Previous studies consistently demonstrate lower use of evidence-based therapies among elderly patients in community practice (2). These practices may, in part, be attributed to the under-representation of older adults (particularly those aged ≥85 years) in clinical trials, leading to lingering uncertainty about the risk-benefit balance of such treatments in these patients (3).
Few studies have compared the effectiveness and safety of drug-eluting stents (DES) with bare-metal stents (BMS) among older patients for whom DES are frequently used “off-label,” and age-related physiological changes may influence treatment outcomes, as well as patient ability to tolerate long-term antiplatelet therapy. Accordingly, we examined data from the National Cardiovascular Data Registry CathPCI (Catheterization Percutaneous Coronary Intervention) Registry linked to Medicare inpatient claims to characterize the treatment and longitudinal outcomes of older PCI patients, particularly those age ≥85 years, and assess the comparative effectiveness of DES versus BMS among each age group.
The CathPCI Registry is the largest U.S. registry of patients undergoing PCI. As previously described (4), trained data abstractors at participating hospitals retrospectively collect detailed clinical information for consecutive PCI patients using standardized data elements and definitions. This study started with all PCI patients ≥65 years of age in the CathPCI Registry from January 2004 to December 2008. International Classification of Diseases-Ninth Revision procedure codes were used to identify index PCI procedures in the Medicare files, which were then linked to CathPCI records using a combination of indirect identifiers (5). The final study population included 471,006 patients from 947 sites successfully matched to Medicare fee-for-service data (Fig. 1).
Death, revascularization, and bleeding were defined both during the index PCI hospitalization (using CathPCI data) and post-discharge (using Medicare data). The International Classification of Diseases-Ninth Revision codes used to identify events were: major bleeding (430 to 432, 578.X, 719.1X, 423.0, 599.7, 626.2, 626.6, 626.8, 627.0, 627.1, 786.3, 784.7, or 459.0), revascularization (36.00, 36.06, 36.07, 36.09, or 36.10 to 36.19), and myocardial infarction (MI) rehospitalization (410.X1). Acute PCI was defined as PCI for ST-segment elevation MI, non–ST-segment elevation MI, or unstable angina.
Baseline characteristics and periprocedural treatments were categorized according to age (65 to 74, 75 to 84, and ≥85 years), PCI setting (acute vs. elective), and stent type (DES versus BMS) and were compared using chi-square tests for categorical variables and Wilcoxon rank sum or Kruskal-Wallis tests for continuous variables.
The cumulative incidence rates for time-to-event outcomes were estimated using Gray's method. For adjusted analyses comparing DES with BMS, a propensity score model was created within each age group comparing DES versus BMS and conditioned on 96 covariates obtained from linked clinical and claims data (6). These covariates included the Charlson index and recent bleeding hospitalization, factors that play a significant role in stent selection among older adults. The propensity score logistic regression models had c-indexes of 0.741 for ages 65 to 74 years, 0.739 for ages 75 to 84 years, and 0.747 for age ≥85 years. The Greedy 5→1 Digit Matching Algorithm was used to match each pair of device types based on the propensity scores. After matching, the distribution of estimated propensity scores for DES patients closely matched that for BMS patients. Adjusted hazard ratios (HRs) comparing DES with BMS were calculated among matched pairs. An inverse probability–weighted method was used as a secondary approach to calculate adjusted HRs.
Baseline characteristics and treatment
Among 471,006 patients age ≥65 years undergoing PCI in our study cohort, 42,154 (9%) were age ≥85 years, 187,656 (40%) were age 75 to 84 years, and 241,196 (51%) were age 65 to 74 years. From 2005 to 2008, the proportion of patients age ≥85 years increased from 7% to 9% among those undergoing elective PCI and from 10% to 13% among those undergoing acute PCI (ptrend <0.001 for both). With increasing age, elective and acute PCI patients were more likely to be female and white, have a higher prevalence of prior heart failure and stroke, and have higher Charlson index scores (Table 1).
Patients age ≥85 years were more likely to undergo multivessel PCI than younger patients (Table 2). Use of DES decreased with age for both elective and acute PCIs. Although lesion complexity increased with age, the use of more complex PCI techniques (e.g., atherectomy) did not differ significantly between age groups. Periprocedural glycoprotein IIb/IIIa inhibitor use decreased with age. Bivalirudin and low-molecular-weight heparin use increased with age among elective PCI patients.
Outcomes across age groups
In-hospital mortality rates were significantly higher among patients age ≥85 years compared with younger patients. The mean duration of follow-up post-discharge was 640.8 ± 423.5 days. Patients ≥85 years of age had the highest long-term mortality among any group, regardless of PCI setting (Fig. 2). Higher rates of MI rehospitalization and bleeding occurred among patients age ≥85 years undergoing acute PCI. Repeat revascularization rates were lowest among patients age ≥85 years treated with acute PCI. These associations persisted after multivariable adjustment (Fig. 3).
Comparison of DES and BMS
Between 2005 and 2008, rates of DES use dropped across all age groups, but the largest decline occurred among patients age ≥85 years (Fig. 4). During elective PCI, DES were less frequently used among patients with prior coronary artery bypass graft or stroke (Table 3). Regardless of PCI indication, DES were more commonly used among patients with low Charlson index scores and those undergoing multivessel PCI (Table 4). Risk-adjusted mortality was lower among DES patients than among BMS patients; however, this difference narrowed with increasing age (pinteraction<0.001) (Table 5). Risk of MI rehospitalization was also significantly lower among DES patients than among BMS patients, with greater risk difference observed with increasing age (pinteraction<0.001). Use of DES was not associated with lower revascularization risk among patients age ≥85 years. Bleeding risk was similar between DES and BMS patients across all age groups.
Several insights emerge from this study, which represents, to the best of our knowledge, the largest report of PCI patients age ≥85 years to date. First, among older patients, an increasing proportion of PCIs are performed in those ≥85 years of age. Second, DES use declined significantly since 2005, with the largest decrease among those age ≥85 years. Third, compared with BMS use, DES use was associated with lower mortality and MI risks across all older age groups without significant difference in repeat revascularization risk.
Although surgical revascularization offers similar or better outcomes compared with PCI (7,8), concerns regarding physical and neurocognitive recovery limit its use among many older adults (9). We observed a greater uptake of newer periprocedural anticoagulant agents that minimize bleeding risk among older patients than younger patients. Thus, the increasing proportion of PCI patients age ≥85 years shown in this study may reflect both an aging population as well as an increased willingness of providers to consider PCI given the contemporary low rates of periprocedural complications (10).
Pivotal DES trials have found a reduction in the need for repeat revascularization but no significant differences in mortality and MI risks compared with BMS (11,12). Nevertheless, these trials enrolled very few elderly patients (<2% age >85 years), and no dedicated, randomized DES-to-BMS comparison has been performed among those aged ≥65 years. In the observational setting, however, use of DES has been associated with lower risks of mortality and MI (6,13,14). Our study expands on the existing literature by focusing on patients ≥85 years of age, adjusting for a comprehensive list of variables relevant to stent selection among older adults, and using several adjustment methods to provide consistent results. In contrast to previous studies (6,13), DES use was not associated with lower repeat revascularization risk among older adults, which may in part be driven by a less invasive attitude in this patient population. However, DES in patients age ≥85 years were associated with a lower risk of MI rehospitalization compared with younger patients.
With reports highlighting the risks of DES thrombosis (15), we observed a decline in DES use, with the largest drop occurring among patients age ≥85 years. The possibility that DES may confer a greater clinical benefit for this high-risk population should be considered. This finding may reflect an outcome difference not observed in randomized trials simply as a function of inadequate statistical power or may represent true differences in DES performance in a community population for whom more than two-thirds of DES implantations were for nonapproved indications (14). Treatment goals for older adults differ from those of younger patients. Reducing repeat revascularization risk may be less of a priority whereas the desire to maximize quality of life by avoiding MI rehospitalization may take precedence for patients age ≥85 years. As such, our study supports consideration of DES use among eligible older adults. The recognition that older adults comprise a growing proportion of PCI patients encourages interest in trials pre-specifying an adequately powered population of enrolled elderly patients to help correct knowledge deficits in this population and incorporating outcomes of relevance (e.g., quality of life).
Given the observational design, associations between treatments and outcomes cannot prove causality. Despite rigorous multivariable adjustment, residual unmeasured factors that influence stent selection and outcomes may skew the benefit toward DES use and cannot be fully accounted for with any adjustment methods. The use of claims data for event classification may result in under-reporting or misclassification of risks; however, these errors should be equally distributed between groups. Angiographic data were reported without central core laboratory adjudication. Finally, we do not have information on longitudinal medication use; this information is key to preventing long-term adverse outcomes.
Patients age ≥85 years are a growing subset of patients undergoing PCI in the United States. In contrast to pivotal stent studies, there was no difference in repeat revascularization between DES and BMS patients age ≥85 years in our study. Compared with BMS use, DES use was associated with lower risks of MI rehospitalization among older PCI patients, particularly among those ≥85 years of age, yet DES use declined the most in this age group. These results emphasize the need to further compare the effectiveness and safety of PCI treatments among elderly patients, focusing on endpoints relevant to this population.
The authors thank Erin LoFrese for her editorial assistance with the manuscript. Ms. LoFrese did not receive compensation for her assistance, apart from her employment at the Duke Clinical Research Institution, where this study was conducted.
This project was sponsored by the Agency for Healthcare Research and Quality, U.S. Department of Health and Human Services, as part of the Cardiovascular Consortium (Project ID: 24-EHC-1, Work Assignment Number: HHSAA290-2005-0032–TO4-WA1, Developing Evidence to Inform Decisions about Effectiveness program). The authors of this paper are responsible for its content, and statements in the paper should not be construed as endorsement by the Agency for Healthcare Research and Quality or the Department of Health and Human Services. Additional support was obtained from the American College of Cardiology National Cardiovascular Data Registry. Dr. Wang has received research grants (to the Duke Clinical Research Institute) from Bristol-Myers Squibb/Sanofi Aventis Partnership, Schering-Plough/Merck & Co., Inc., The Medicines Company, Heartscape Technologies, Inc., Canyon Pharmaceuticals, and Eli Lilly and Company/Daiichi Sankyo alliance; and consulting or honoraria from Medco Health Solutions, Inc., and AstraZeneca. Dr. Masoudi has received salary support from the American College of Cardiology; is senior medical officer for the National Cardiovascular Data Registry (significant); and is a consultant for the Oklahoma Foundation for Medical Quality. Dr. Anstrom has received research and salary support from Alexion Pharmaceuticals, AstraZeneca, Bristol-Myers Squibb, Eli Lilly and Company, Innocoll Pharmaceuticals, Medtronic, Pfizer Inc., and Proctor & Gamble (modest); serves on the data safety monitoring boards of Pfizer Inc. and Vertex Pharmaceuticals (modest); and provides consulting services to Pacific Therapeutics Ltd., Bristol-Myers Squibb, and AstraZeneca (modest). Dr. Peterson has received research support from Bristol-Myers Squibb/Sanofi Aventis Partnership, Merck & Co., Inc., Johnson & Johnson, and Eli Lilly and Company (all significant). Dr. Rumsfeld is the chief science officer for the National Cardiovascular Data Registry. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- bare metal stent(s)
- drug-eluting stent(s)
- hazard ratio
- myocardial infarction
- percutaneous coronary intervention
- Received August 1, 2011.
- Revision received September 20, 2011.
- Accepted October 11, 2011.
- American College of Cardiology Foundation
- US Census Bureau
- Alexander K.P.,
- Roe M.T.,
- Chen A.Y.,
- et al.
- Brindis R.G.,
- Fitzgerald S.,
- Anderson H.V.,
- et al.
- Douglas P.S.,
- Brennan J.M.,
- Anstrom K.J.,
- et al.
- Groeneveld P.W.,
- Matta M.A.,
- Greenhut A.P.,
- et al.