Redo Coronary Artery Bypass Grafting: The Hidden Challenges of Propensity-Matched Comparisons

All articles of this category (opens in new window) Letter to: Long-Term Outcomes after Redo Coronary Artery Bypass Grafting: A Propensity-Matched Analysis of On-Pump and Off-Pump Techniques Thorac Cardiovasc Surg eFirst DOI: 10.1055/a-2900-0233 10.1055/a-2900-0233 We read with great interest the article by Braun et al examining long-term outcomes after redo coronary artery bypass grafting (CABG) using propensity-matched comparisons of on-pump and off-pump techniques.[ 1 ] The authors should be commended for addressing a challenging and relatively understudied area of coronary surgery. Nevertheless, we believe several methodological issues deserve consideration when interpreting the reported equivalence in long-term outcomes. Although propensity score matching reduces measured baseline differences, its validity depends on the selection of clinically relevant covariates. In redo CABG, several important determinants of operative strategy and prognosis appear not to have been incorporated into the matching model. These include the presence of a patent left internal mammary artery or other functioning grafts, the indication for reoperation (graft failure vs. progression of native coronary disease), the interval from the primary CABG, coronary anatomical complexity and target-vessel quality, completeness of planned revascularization, severe ascending aortic calcification or porcelain aorta, and surgeon experience with off-pump revascularization.[ 2 ] [ 3 ] Each of these factors may substantially influence both the choice of surgical technique and subsequent outcomes. An additional finding deserving further discussion is that the on-pump cohort underwent a greater number of distal anastomoses, while long-term survival remained comparable between groups. Without information regarding the completeness of revascularization or residual ischemic burden, it is difficult to determine whether this reflects differences in coronary anatomy, the presence of patent previous grafts, or selection of technically less complex patients for the off-pump approach. Consequently, residual confounding by indication cannot be excluded despite propensity matching. Redo CABG represents one of the most individualized procedures in cardiac surgery, with operative decisions frequently influenced by anatomical and technical considerations that are difficult to capture in retrospective databases. We therefore suggest that the findings should be interpreted as demonstrating comparable outcomes among carefully selected patients rather than definitive evidence of equivalence between the two techniques.[ 2 ] [ 3 ] Future multicenter studies incorporating detailed anatomical, graft-related, and surgeon-specific variables would further strengthen the evidence guiding the optimal surgical strategy for redo coronary revascularization. Future studies incorporating detailed anatomical, graft-related, and surgeon-specific variables are needed to better define the optimal strategy for redo coronary revascularization. In addition, artificial intelligence–based approaches may offer new opportunities to integrate complex clinical and imaging data, potentially improving individualized risk prediction and surgical decision-making in this challenging population.[ 4 ] Publication History Received: 21 July 2026 Accepted: 06 August 2026 Article published online: 12 August 2026 © 2026. Thieme. All rights reserved. Georg Thieme Verlag KG Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany