Author + information
- Giuseppe Tarantini, MD, PhD∗ (, )
- Gianpiero D’Amico, MD,
- Sorin J. Brener, MD,
- Paola Tellaroli, MSc, PhD and
- Gregg W. Stone, MD
- ↵∗Ospedale di Padova, Centro Cardiologico Gallucci, Clinica Cardiologica, Via Giustiniani 2, Padova 35128, Italy
We thank Dr. Tamis-Holland and colleagues for their interest in our article (1).
We recognize that combining prospective observational studies with randomized controlled trials (RCTs) might be a limitation, but this was done to increase the statistical power of our observations. We acknowledge that this method might increase heterogeneity; thus, we performed a subgroup meta-analysis to explore the impact of the study type (i.e., retrospective vs. prospective) as a potential source of heterogeneity, as suggested by the Cochran Handbook (2). Moreover, although it might be interesting to group the studies as retrospective, observational prospective, and RCTs, we could not proceed in this way because of the low number of prospective studies included in the meta-analysis and the fact that at least 4 studies need to be included in each subgroup for meaningful results (3). In any case, we failed to find any heterogeneity in the subgroup of prospective studies. Finally, there are no prospective observational studies comparing infarct-related artery (IRA) percutaneous coronary intervention (PCI) versus staged PCI and single-stage multivessel (MV) PCI versus staged PCI; the study by Kornowski et al. (4) is a post hoc retrospective analysis from the HORIZONS-AMI (Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction) trial.
We regret that the commentators did not note that we did perform a Bayesian analysis, besides the frequentist method, and it fully confirmed the latter. Applying the suggested cross-design meta-analysis, the significant impact of the IRA-only versus single-stage MV-PCI on long-term mortality is lost (Figure 1). However, a major drawback of this approach remains the low numbers of studies included in each group.
In the CvLPRIT trial (Complete Versus Lesion-Only Primary PCI Trial) (5), the choice of the strategy was not randomized, and none of the analyses used “as treated” methodology. Notwithstanding, the complete revascularization was recommended and performed during the index procedure in most patients, and thus, we decided to label this trial as single-stage MV-PCI strategy. As shown in Figure 2, reclassifying the latter trial as staged MV-PCI or removing it from the analysis (as suggested in the letter) does not change significantly the results of our analysis (p = 0.005 and p = 0.0144, respectively).
We thank the commentators again and agree with the need for prospective, randomized clinical trials addressing this important clinical dilemma.
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
- Tarantini G.,
- D’Amico G.,
- Brener S.J.,
- et al.
- ↵Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0. Updated March 2011. The Cochrane Collaboration. Available at http://handbook.cochrane.org. Accessed November 10, 2016.
- Fu R.,
- Gartlehner G.,
- Grant M.,
- et al.
- Kornowski R.,
- Mehran R.,
- Dangas G.,
- et al.
- Gershlick A.H.,
- Khan J.N.,
- Kelly D.J.,
- et al.