T he need for emergency cardiac surgery has decreased dramatically from 6% to 10%, 1 during the era of balloon angioplasty, to 0.1% to 0.4% in the current era of stents because of the many advances in technology, techniques, adjunctive pharmacotherapy, and operator experience. [2][3][4] Despite this progress, concerns remain about performing percutaneous coronary intervention (PCI) at centers without on-site surgical backup, especially regarding nonprimary PCI for conditions other than ST-segment-elevation myocardial infarction (STEMI). Because primary PCI confers longer survival and timely reperfusion, increased access to primary PCI was encouraged. Subsequently, numerous studies showed that safety and efficacy of primary PCI are similar in centers with and without on-site surgical capability. 4,5 Current American College of Cardiology Foundation/ American Heart Association/Society for Cardiovascular Angiography and Interventions guidelines recommend that primary PCI for STEMI be performed at centers without on-site surgical backup (class IIa, level of evidence: B).
1,5
Editorial see p 365 Clinical Perspective on p 401Conversely, nonprimary PCI has been a major issue in this debate; no survival benefit supports allowing nonprimary PCI at centers without on-site surgical backup. Nonprimary Background-Emergency coronary artery bypass grafting for unsuccessful percutaneous coronary intervention (PCI) is now rare. We aimed to evaluate the current safety and outcomes of primary PCI and nonprimary PCI at centers with and without on-site surgical backup. Methods and Results-We performed an updated systematic review and meta-analysis by using mixed-effects models. We included 23 high-quality studies that compared clinical outcomes and complication rates of 1 101 123 patients after PCI at centers with or without on-site surgery. For primary PCI for ST-segment-elevation myocardial infarction (133 574 patients), all-cause mortality (without on-site surgery versus with on-site surgery: observed rates, 4.8% versus 7.2%; pooled odds ratio [OR], 0.99; 95% confidence interval, 0.91-1.07; P=0.729; I 2 =3.4%) or emergency coronary artery bypass grafting rates (observed rates, 1.5% versus 2.4%; pooled OR, 0.76; 95% confidence interval, 0.56-1.01; P=0.062; I 2 =42.5%) did not differ by presence of on-site surgery. For nonprimary PCI (967 549 patients), all-cause mortality (observed rates, 1.6% versus 2.1%; pooled OR, 1.15; 95% confidence interval, 0.94-1.41; P=0.172; I 2 =67.5%) and emergency coronary artery bypass grafting rates (observed rates, 0.5% versus 0.8%; pooled OR, 1.14; 95% confidence interval, 0.62-2.13; P=0.669; I 2 =81.7%) were not significantly different. PCI complication rates (cardiogenic shock, stroke, aortic dissection, tamponade, recurrent infarction) also did not differ by on-site surgical capability. Cumulative meta-analysis of nonprimary PCI showed a temporal decrease of the effect size (OR) for all-cause mortality after 2007. Conclusions-Clinical outcomes and complication rates of PCI at centers without on-...