Mark A. Hlatky, MD; Derek B. Boothroyd, PhD; Laurence Baker, PhD; Dhruv S. Kazi, MD, MS; Matthew D. Solomon, MD, PhD; Tara I. Chang, MD, MS; David Shilane, PhD; Alan S. Go, MD
This article was published at www.annals.org on 23 April 2013.
Grant Support: By grant HL099872 from the National Heart, Lung, and Blood Institute.
Potential Conflicts of Interest: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M12-1564.
Reproducible Research Statement: Study protocol: The protocol is fully described in the manuscript. Statistical code: Available on request. Data set: Medicare data sets are available to qualified researchers but cannot be released by the investigators.
Requests for Single Reprints: Mark A. Hltaky, MD, Stanford University School of Medicine, HRP Redwood Building, 259 Campus Drive, Stanford, CA 94305-5405; e-mail, email@example.com.
Current Author Addresses: Drs. Hlatky, Boothroyd, Baker, and Shilane: Stanford University School of Medicine, HRP Redwood Building, 259 Campus Drive, Stanford, CA 94305-5405.
Dr. Kazi: UCSF Division of Cardiology, San Francisco General Hospital, 1001 Potrero Avenue, SG1, San Francisco, CA 94110.
Drs. Solomon and Go: Division of Research, Kaiser Permanente Northern California, 2000 Broadway, Oakland, CA 94612.
Dr. Chang: Stanford University School of Medicine, Nephrology Division, 777 Welch Road, Room D100, Palo Alto, CA 94304.
Author Contributions: Conception and design: M.A. Hlatky, L. Baker, A.S. Go.
Analysis and interpretation of the data: M.A. Hlatky, D.B. Boothroyd, D.S. Kazi, M.D. Solomon, T.I. Chang, D. Shilane, A.S. Go.
Drafting of the article: M.A. Hlatky.
Critical revision of the article for important intellectual content: D.B. Boothroyd, D.S. Kazi, M.D. Solomon, T.I. Chang, D. Shilane, A.S. Go.
Final approval of the article: M.A. Hlatky, D.B. Boothroyd, L. Baker, D.S. Kazi, M.D. Solomon, T.I. Chang, D. Shilane, A.S. Go.
Provision of study materials or patients: L. Baker.
Statistical expertise: M.A. Hlatky, D.B. Boothroyd, D. Shilane.
Obtaining of funding: M.A. Hlatky, A.S. Go.
Administrative, technical, or logistic support: M.A. Hlatky, L. Baker.
Collection and assembly of data: D.B. Boothroyd, L. Baker.
Hlatky M., Boothroyd D., Baker L., Kazi D., Solomon M., Chang T., Shilane D., Go A.; Comparative Effectiveness of Multivessel Coronary Bypass Surgery and Multivessel Percutaneous Coronary Intervention: A Cohort Study. Ann Intern Med. 2013;158:727-734. doi: 10.7326/0003-4819-158-10-201305210-00639
Download citation file:
Published: Ann Intern Med. 2013;158(10):727-734.
Randomized trials of coronary artery bypass graft (CABG) surgery and percutaneous coronary intervention (PCI) suggest that patient characteristics modify the effect of treatment on mortality.
To assess whether clinical characteristics modify the comparative effectiveness of CABG versus PCI in an unselected, general patient population.
Observational treatment comparison using propensity score matching and Cox proportional hazards models.
United States, 1992 to 2008.
Medicare beneficiaries aged 66 years or older.
Multivessel CABG or multivessel PCI.
The CABG–PCI hazard ratio (HR) for all-cause mortality, with prespecified treatment-by-covariate interaction tests, and the absolute difference in life-years of survival in clinical subgroups after CABG or PCI, both over 5 years of follow-up.
Among 105 156 propensity score–matched patients, CABG was associated with lower mortality than PCI (HR, 0.92 [95% CI, 0.90 to 0.95]; P < 0.001). Association of CABG with lower mortality was significantly greater (interaction P ≤ 0.002 for each) among patients with diabetes (HR, 0.88), a history of tobacco use (HR, 0.82), heart failure (HR, 0.84), and peripheral arterial disease (HR, 0.85). The overall predicted difference in survival between CABG and PCI treatment over 5 years was 0.053 life-years (range, −0.017 to 0.579 life-years). Patients with diabetes, heart failure, peripheral arterial disease, or tobacco use had the largest predicted differences in survival after CABG, whereas those with none of these factors had slightly better survival after PCI.
Treatments were chosen by patients and physicians rather than being randomly assigned.
Multivessel CABG is associated with lower long-term mortality than multivessel PCI in the community setting. This association is substantially modified by patient characteristics, with improvement in survival concentrated among patients with diabetes, tobacco use, heart failure, or peripheral arterial disease.
National Heart, Lung, and Blood Institute.
VINOD K CHAUBEY
ST VINCENT HOSPITAL, UNIVERSITY OF MASSACHUSETTS
April 30, 2013
Did author consider classifying the PCI into balloon angioplasty vs. bare metal stents vs. drug eluting stents for mortality end point comparison?
To the editor: Although quite informative study, authors did not compare the rates of revascularization in PCI versus CABG. Recent review including meta- analysis showed that patients who has high risk (diabetes mellitus, multiple vessel disease) there is no benefit in terms of cardiac death and myocardial infarction in comparisons with various stents ( drug vs. bare metal stents) technique. In addition, what is the rate of repeat revascularization in patient who underwent CABG was not mentioned either which is prudent in patient who has multiple risk factors? Also, it would also been worthwhile to classify patients who underwent PCI into the various type of PCI: balloon angioplasty vs. bare metal stent vs. drug eluting stent if any and then to look for mortality end points.
Junji Kumasawa, MD , Noriaki Kurita, MD , Shunichi Fukuhara, MD, MSc, DMSc
Department of Healthcare Epidemiology, Kyoto University Graduate School of Medicine and Public Health, Japan
June 18, 2013
Propensity score analysis might not reduce strong confounding by indication bias.
To the editor. Although the authors should be commended for their attempts to assess the clinical characteristics that could modify the comparative effectiveness of CABG versus PCI in an unselected general population, the limitation of propensity score analysis should receive a greater emphasis when there are several unmeasured confounding factors. As the authors mentioned, medication and smoking status are critical confounding factors, but there are several other unmeasured critical confounding factors, such as severe aortic calcification, immunosuppression, and systemic infection. McNulty et al. explicitly identified these factors contributed to the ineligibility for CABG in more than 50% of patients undergoing PCI for left main coronary artery disease; and ineligible patients have a five-fold mortality risk compared with those who are eligible1). Thus, the association of reduced death risk with CABG versus PCI would be overestimated by such residual confounding factors. Stukel et al. reported that propensity score analysis could not reduce bias due to unmeasured confounding factors when such confounding was strong. They estimated the effectiveness of invasive cardiac management on mortality of patients with acute myocardial infarction and demonstrated that the reduction in the risk of death by using propensity score analysis was greater than that using instrumental variable method 2). If possible, the authors should estimate average treatment effectiveness using the instrumental variable method as a sensitivity analysis. The regional CABG performing rate can be used as an instrumental variable, as the regional cardiac catheterization rate was used in Stukel’s study.
1. McNulty EJ, Ng W, Spertus JA, Zaroff JG, Yeh RW, Ren XM, et al. Surgical candidacy and selection biases in nonemergent left main stenting: implications for observational studies. JACC Cardiovasc Interv. 2011;4(9):1020-7.
2. Stukel TA, Fisher ES, Wennberg DE, Alter DA, Gottlieb DJ, Vermeulen MJ. Analysis of observational studies in the presence of treatment selection bias: effects of invasive cardiac management on AMI survival using propensity score and instrumental variable methods. Jama. 2007;297(3):278-85.
Mark Hlatky, MD, Derek Boothroyd, PhD
July 30, 2013
We thank Kumasawa and colleagues for their interest in our study and comments. Propensity score analysis can only adjust for measured confounding factors, and we agree that unmeasured confounders can affect estimates of treatment effectiveness. However, a major focus of our study was to demonstrate heterogeneity of treatment effectiveness due to key clinical factors, and how this heterogeneity could lead to greatly different changes in mortality. It is quite unlikely that the highly significant variation in treatment effectiveness that we found related to measured factors can be entirely explained by imbalances in unmeasured factors. It is also important to note that heterogeneity of treatment effects were also seen in analysis of data from randomized trials (1), which balance both measured and unmeasured factors.In principle, instrumental variables can control for unmeasured confounders in the assessment of main effects, but they cannot necessarily assess heterogeneity of treatment effects.
Mark Hlatky MD
Derek Boothroyd PhD
1. Hlatky MA, Boothroyd DB, Bravata DM, Boersma E, Booth J, BrooksMM, et al. Coronary artery bypass surgery compared with percutaneous coronaryinterventions for multivessel disease: a collaborative analysis of individual patientdata from ten randomised trials. Lancet. 2009;373:1190-7.
to gain full access to the content and tools.
Learn more about subscription options.
Register Now for a free account.
Cardiology, Healthcare Delivery and Policy, Coronary Heart Disease, Percutaneous Coronary Intervention.
Results provided by:
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only