Safety and Efficacy of Transcatheter Aortic Valve Implantation at Centres With Versus Without On-Site Cardiac Surgery: A Systematic Review and Meta-Analysis

BACKGROUND Aortic stenosis (AS) is the most common heart valve disorder causing significant complications. While transcatheter aortic valve implantation (TAVI) offers a less invasive alternative to surgery, debate remains about whether it should require on-site cardiac surgery backup. This systematic review and meta-analysis evaluate existing literature comparing outcomes of TAVI performed at centres with versus without on-site cardiac surgery, aiming to guide optimal patient management and institutional decision-making. METHODS A comprehensive literature search was conducted in multiple databases up to May 2025 to retrieve relevant studies. Random-effects meta-analyses pooled risk ratios (RR) or mean differences. Because definitions of surgical backup varied across studies, backup organisation was extracted for each study. Risk of bias was assessed with the Newcastle-Ottawa Scale and certainty of evidence with the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system. RESULTS Six observational cohort studies (n=6,109 patients) were included. No statistically significant differences were observed for in-hospital mortality (four studies; n=3,463; RR= 0.94), 30-day mortality (five studies; n=4,880; RR= 0.87), or emergent conversion to open-heart surgery (four studies; n=4,136; RR= 0.78). Most secondary outcomes were likewise not statistically different, but several had substantial heterogeneity or unstable leave-one-out results. In one matched registry study from the included studies, in-hospital mortality among patients requiring emergent cardiac surgery after intraprocedural complications was high in both settings (50% vs 62.5%). CONCLUSIONS This observational meta-analysis did not detect statistically significant differences in short-term mortality, emergent conversion, or major complications between centres with and without on-site cardiac surgery. However, the absence of statistically significant differences does not establish safety equivalence. Certainty of evidence was low to very low, and findings should be interpreted cautiously in light of residual confounding, publication bias, and variation in backup models. Accordingly, future high-quality studies are needed to validate these findings.