Is It Safe to Perform PCI Without a Safety Net?

Journal Wrap | A recent Veterans Affairs study comparing percutaneous coronary intervention (PCI) centers with and without on-site cardiothoracic surgical surgery capabilities demonstrated that allowing PCI at centers without surgical backup is not only safe, but can improve patient outcomes.

The need for subsequent revascularization, however, was higher at centers without surgical backup—particularly with repeat PCI.

Thomas Maddox, MD, MSc, and co-authors looked at data from 24,387 patients (mean age = 63 years; 98.4% male) who underwent procedures at 59 VA facilities from October 2007 to September 2010; 27.1% of patients (6,616) underwent procedures at 18 facilities without on-site surgical backup. They analyzed patient access (determined by patient drive time), procedural complications, 1-year mortality, myocardial infarction (MI), and rates of subsequent revascularization by facility. Results were stratified by procedural indication (ST-segment elevation MI [STEMI] vs. NSTEMI) and PCI volume.

When PCI facilities without on-site surgical backup were included in the scenario, median patient drive time to the nearest PCI facility was shortened by 90.8 minutes (p < 0.001). The reduction was even greater for patients who were undergoing primary PCI for STEMI (96.1 minutes), versus those with NSTEMI/unstable angina (91.4 minutes) and those having elective procedures (87.2 minutes).

Importantly, that improvement in patient access was not accompanied by any signals of harm: rates of emergent coronary artery bypass grafting during the procedure and procedural mortality rates were 0.1% or lower (whether or not the center had on-site surgical backup).

In terms of the primary outcome (adjusted 1-year all-cause mortality or MI hospitalization), rates were similar between facilities with and without on-site surgical backup (HR in PCI facilities without relative to those with on-site surgery = 1.02; 95% CI 0.87-1.2). This interaction was not modified by either PCI indication or PCI volume.

However, at sites without on-site surgical backup, there was a greater risk of subsequent unplanned revascularization in the year after index PCI (15.2% vs. 12.7%; adjusted HR = 1.21; 95% CI 1.03-1.42). This association was mainly driven by a higher risk of repeat PCI (adjusted HR = 1.28; 95% CI 1.07-1.52).

"The reasons behind this finding are unclear, but may represent a lower initial PCI success rate and/or a more conservative approach by interventionalists at these facilities," Dr. Maddox and colleagues wrote. "Regardless of the reasons, this difference in revascularization rates was not accompanied by a safety signal of harm at 1 year."

Ultimately, the authors concluded, providing PCI facilities without on-site cardiothoracic surgery in an integrated health care system, then, does not comprise procedural or 1-year outcomes. "Current PCI guidelines allow for the procedure at facilities without on-site cardiothoracic surgery, but call for appropriate program development and quality oversight," the authors wrote, noting the higher rate of unplanned revascularization when on-site backup was not available.

As Maddox et al. pointed out, the VA health care system has employed this strategy—with strict quality oversight—since 2005. Previous clinical trials have demonstrated the safety and effectiveness of PCI without the "safety net" of on-site cardiothoracic surgical backup, and the findings from this real-world comparative-effectiveness analysis match those results.

However, the investigators acknowledge that it is too soon to apply these findings to female patients (given that they comprised only 1.6% of the study population), or to settings outside of the VA. The quality oversight features "inherent in the VA CART program can be replicated in non-VA settings, and the national move towards greater health care integration and accountable care organizations may necessitate emulating the VA's approach."


  1. Maddox TM, Stanislawski MA, O'Donnell C, et al. Circulation. 2014 September 4. [Epub ahead of print]

< Back to Listings