Issues Needing Clarification Regarding Population Characteristics in the Analysis From the National Cardiovascular Data Registry
Author + information
- Mikail Yarlioglues, MD,
- Mahmut Akpek, MD and
- Mehmet G. Kaya, MD⁎ (drmgkaya{at}yahoo.com)
- ↵⁎Department of Cardiology, Erciyes University School of Medicine, 38039 Melikgazi, Kayseri, Turkey
We read with interest the paper by Baklanov et al. (1), but we think that some issues need to be clarified regarding the population characteristics of the Prevalence and Outcomes of Transradial Percutaneous Coronary Intervention for ST-Segment Elevation Myocardial Infarction: Analysis From the National Cardiovascular Data Registry (2007 to 2011).
First, there is no information about sheath sizes used in the study groups. One of the endpoints was bleeding, and overt access site bleeding and retroperitoneal hemorrhage within 72 h of percutaneous coronary intervention (PCI) were included in the definition of bleeding. Doyle et al. (2) reported that using sheath sizes >6-F was identified as a strong independent predictor of major femoral bleeding. Trimarchi et al. (3) reported that adopting a sheath size of 8-F or larger was independently related to independent predictors of retroperitoneal hematoma after PCI.
Second, the incidence of high-risk C-type coronary lesions was significantly greater in the femoral access group compared with the radial access group. Barbash et al. (4) concluded that a type-C coronary lesion was an angiographic predictor of PCI failure for ST-segment elevation myocardial infarction. In addition, Wilensky et al. (5) reported that PCI for complex lesions was associated with increased in-hospital mortality. When considering that PCI success and mortality were defined as primary endpoints of the study, results may be affected because of this difference between groups.
Finally, patients with renal failure receiving hemodialysis were significantly more common in the femoral access group. It is well-known that patients with renal failure have bleeding disorders due to uremic platelet dysfunction (6). Skin bleeding time is used for the evaluation of platelet dysfunction (7). Dialysis may partially correct these defects, but cannot totally eliminate them. The hemodialysis process itself may in fact contribute to bleeding. Baklanov et al. (1) did not clarify whether they evaluated bleeding disorders using skin bleeding time and excluded this type of patient from the study population. They used propensity score (PS) matching in the study. Even though the PS can balance observed baseline covariates between groups, they do nothing to balance unmeasured characteristics and confounders. PS analyses have the limitation that remaining unmeasured confounding may still be present. According to their paper, the PS did not include skin bleeding time or any different method to identify the presence of a bleeding disorder. Therefore, this situation may have affected the results of the study.
- American College of Cardiology Foundation
