Author + information
- Antonio Colombo, MD∗ (, )
- Alaide Chieffo, MD,
- SECURITY Investigators
- ↵∗Interventional Cardiology Unit, San Raffaele Scientific Institute, Milan 20145, Italy
We thank Dr. Kaneda and Dr. Perneger for their interest in our paper (1), giving us the possibility to clarify some relevant aspects of the statistical analysis of our SECURITY (Second-Generation Drug-Eluting Stent Implantation Followed by 6- Versus 12-Month Dual Antiplatelet Therapy) randomized noninferiority trial.
We submitted both letters to Professor Bruno Mario Cesana, the statistician of the SECURITY trial, and his answer is the following:
Because the confidence interval (CI) calculation proposed methods (2,3) for the difference between 2 proportions give different results possibly leading to different conclusions in the analysis of noninferiority studies in which the CI approach is used, we calculated, according to Thomas and Gart (4), the CI limits of the odds ratio (OR) that can be “translated” into the upper and the lower CI limits for the difference between 2 proportions. We used the approximate OR 90% CI, due to the much higher coverage of the “exact” OR CI.
As a result, the data in Table 5 give an OR of 1.21 with 90% CI: 0.78 to 1.89, translated as -0.0093 to 0.024. Consequently, “6 to 12 months” should be considered “6 months” and “95% CI” as “90% CI” only for the primary efficacy composite endpoint. In addition, the 90% CI are 0.49% to 1.29% for the comparison at 6 months and -0.93% to 2.4% for the comparison at 12 months.
The unilateral 95% upper limit of 2.4%, greater than 2%, does not allow us to conclude for the noninferiority. However, from the results of the landmark analysis (not shown in the paper because of space limitations) after having excluded the events that occurred in the first 6 months in which the treatment is equal for the 2 arms and in which, by chance, more events occurred in the 6-month dual antiplatelet therapy (DAPT) arm, there are 23 events on 674 patients (0.0341; 95% CI: 0.0234 to 0.0480) in the 6-month DAPT arm against 22 events on 712 patients (0.0341; 95% CI: 0.0210 to 0.0438) in the 12-month DAPT arm.
As we have been cautious in the SECURITY paper (1) to state the conclusiveness of noninferiority, the same applies to the lack of noninferiority. The bottom line is that with insufficient power, the conclusions remain uncertain. Nevertheless, the data obtained from the SECURITY study are a contribution to test the hypothesis that shorter duration of DAPT following implantation of second-generation drug-eluting stents is as safe and effective as 12 months of therapy. This study underscores that larger patient enrollment is necessary to answer this question.
Please note: The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation