Response to Letters Regarding Article, “Two-Year Clinical Outcomes With Drug-Eluting Stents for Diabetic Patients With De Novo Coronary Lesions: Results From a Real-World Multicenter Registry”
We thank Barlis and colleagues for their legitimate1 query about the ability of the propensity score regression technique to balance covariates between the somewhat heterogeneous groups. Our decision to report the results of a propensity score regression analysis conducted across the overall study population was based on a number of considerations. Nevertheless, application of propensity score matching (as suggested by Barlis and colleagues) in 2 well-balanced but smaller groups (253 patients each) substantially confirms the main results. Indeed, the result for the main outcome measure (major events at 24 months in terms of death/myocardial infarction/target vessel revascularization) for patients treated with drug-eluting stents was practically identical to that of the published analysis: After propensity score matching, the hazard ratio (HR) versus those treated with bare metal stents was 0.77 (95% confidence interval [CI], 0.54 to 1.10); the HR without matching was 0.77 (95% CI, 0.59 to 1.01). The results of the matched analysis were also consistent for death (HR, 0.84; 95% CI, 0.50 to 1.41) and death/myocardial infarction (HR, 0.82; 95% CI, 0.53 to 1.25). The only apparent difference was a loss of statistical significance for target vessel revascularization (HR, 0.70; 95% CI, 0.42 to 1.14 with matching versus 0.66; 95% CI, 0.46 to 0.96 without matching), perhaps because of the more limited numbers.
Relative to the other points raised by Barlis and colleagues, for a cohort study focusing on 2-year clinical outcome we believe that the (inevitable) lack of angiographic follow-up data cannot be considered a weakness, especially given the absence of evidence on the clinical relevance of (or benefits of treating) “silent” restenosis. Indeed, we think that angiographically driven target vessel revascularization could lead to overestimates of the clinical benefits of drug-eluting stents.2 We avoided (in full agreement with the editors of the journal) reporting subgroup comparisons of the different types of drug-eluting stents, which would have been extremely vulnerable to type 2 statistical errors because of the limited numbers (for the record, no statistical trend was apparent). As we acknowledged in the article, the registry data precluded use of the Academic Research Consortium classification of stent thrombosis.3
In reply to Dr Watada, we do recognize in retrospect that the subgroup of patients using insulin therapy (composed of type 1 and type 2 diabetic patients) enrolled in our study should have been termed “insulin requiring” (as distinct from “insulin dependent”) in accordance with the established guidelines.4 We would like to thank Dr Watada for highlighting the relevance of this distinction.
D'Agostino RB, Jr. Propensity scores in cardiovascular research. Circulation. 2007; 115: 2340–2343.
Pinto DS, Stone GW, Ellis SG, Cox DA, Hermiller J, O'Shaughnessy C, Mann JT, Mehran R, Na Y, Turco M, Caputo R, Popma JJ, Cutlip DE, Russell ME, Cohen DJ. Impact of routine angiographic follow-up on the clinical benefits of paclitaxel-eluting stents: results from the TAXUS-IV trial. J Am Coll Cardiol. 2006; 48: 32–36.
Cutlip DE, Windecker S, Mehran R, Boam A, Cohen DJ, van Es GA, Steg PG, Morel MA, Mauri L, Vranckx P, McFadden E, Lansky A, Hamon M, Krucoff MW, Serruys PW. Clinical end points in coronary stent trials: a case for standardized definitions. Circulation. 2007; 115: 2344–2351.
Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care. 1997; 20: 1183–1197.