Emergency coronary artery bypass surgery after PCI at hospitals with versus without on-site surgical backup: a systematic review and meta-analysis
A Omayer, R Ahmed, K Khan, A Sharif, M Shahid, T Shahid, S Basheer, M F Anwar, M U Soomro, R Ali, S M H HoqueAbstract
Background
The safety of percutaneous coronary intervention (PCI) at hospitals without on-site surgical backup (NSOS) remains debated, particularly regarding emergency coronary artery bypass grafting (eCABG). Procedural practice differences may contribute to outcome variability.
Purpose
To compare eCABG after PCI in hospitals with versus without on-site surgical backup (SOS vs NSOS) and to explore whether study-level clinical and procedural factors explain between-study heterogeneity.
Methods
Following PRISMA, comparative SOS versus NSOS PCI studies were identified from PubMed, Web of Science, Scopus, and the Cochrane Library. eCABG outcomes were pooled using a random-effects Mantel–Haenszel model. Procedural characteristics and univariable meta-regression were used to explore heterogeneity. Quality assessment used NOS (Newcastle–Ottawa Scale) for cohorts and RoB 2 for RCTs; meta-regression was performed in RStudio (metafor).
Results
Of 1,300 records screened, 13 comparative studies were included. In the primary analysis, SOS centres demonstrated a non-significant trend toward higher odds of eCABG compared with NSOS (OR = 1.45, 95% CI 0.99–2.13), with substantial heterogeneity (I² = 85%). Sensitivity analysis excluding an influential study yielded a significant association with higher eCABG in SOS centres (OR = 1.75, 95% CI 1.32–2.31) and reduced heterogeneity (I² = 34%). Procedural characteristics were broadly similar, with no difference in overall stent use. Vessel selection was comparable for LAD, RCA, and LCA/LM, while the circumflex artery was targeted slightly more often in NSOS centres (OR = 0.94, 95% CI 0.92–0.95). Stent type differed: bare-metal stent use was lower in SOS centres (OR = 0.74, 95% CI 0.55–0.99; I² = 93%) and drug-eluting stent use was higher (OR = 1.49, 95% CI 1.05–2.11; I² = 99%). Meta-regression suggested heterogeneity was partly related to case mix: higher male and NSTEMI proportions were associated with a larger SOS–NSOS difference, whereas older age and higher prior PCI prevalence were associated with a smaller difference; the number of stents used also influenced effect size. Diabetes, hypertension, smoking, prior CABG, and 1–3-vessel disease did not explain heterogeneity. All RCTs were low risk of bias, while cohort studies scored 6–9 on NOS, with most ≥8 and none rated poor quality.
Conclusions
Across 13 studies, eCABG after PCI showed borderline higher odds in SOS centres in the primary analysis, becoming significant after sensitivity analysis with improved consistency. While overall stent use and most target vessels were comparable, NSOS centres targeted the circumflex slightly more often and SOS centres used fewer bare-metal stents and more drug-eluting stents, indicating procedural practice differences. Study-level case mix and stent utilisation may contribute to heterogeneity in eCABG risk between SOS and NSOS settings.Forest Plot of Emergency CABG After PCIFor image description, please refer to the figure legend and surrounding text.