Randomized Trial to Compare Percutaneous Coronary Intervention Between Massachusetts Hospitals With Cardiac Surgery On-Site and Community Hospitals Without Cardiac Surgery On-Site - MASS COMM


Current American College of Cardiology Foundation (ACCF)/American Heart Association (AHA) guidelines list elective percutaneous coronary intervention (PCI) as a Class IIb indication for institutions without on-site cardiac surgery. Most of this is based on earlier observational data showing worse outcomes in patients undergoing PCI at such institutions. The current trial sought to study this question in a randomized, controlled fashion.


Outcomes after nonemergent PCI at sites without on-site cardiac surgery would be noninferior to PCI at sites with 24/7 cardiac surgery.

Study Design

  • Randomized
  • Blinded
  • Parallel
  • Stratified

Patient Populations:

  • Age >18 years
  • >50% stenosis
  • All target lesions approachable at primary hospital; hospital required approval by Massachusetts Department of Public Health
  • Signed Collaboration Agreement with on-site surgery hospital (24/7 back-up, patient arrival within 60 minutes)
  • >300 PCI/year at that institution, 36 primary PCI
  • Interventionalist performs >75 PCIs/year and is American Board of Internal Medicine certified

    Number of enrollees: 3,691
    Duration of follow-up: 1 year
    Mean patient age: 64.5 years
    Percentage female: 33%
    Ejection fraction: 56%


  • Left ventricular ejection fraction <20%
  • Target lesions with:
    - Unprotected left main stenosis >50%
    - Device other than balloon angioplasty prior to stent
    - Saphenous vein graft location
    - Vessel serving only viable myocardium

Primary Endpoints:

  • Composite of all-cause mortality, MI, repeat coronary revascularization, or stroke (MACE) at 30 days (safety) and 12 months (effectiveness)

Secondary Endpoints:

  • All-cause mortality
  • Repeat revascularization
  • Stroke
  • Ischemia-driven target vessel and target lesion revascularization
  • Academic Research Consortium (ARC) defined definite or probable stent thrombosis
  • Emergency CABG
  • Emergency or urgent PCI
  • Major vascular complications

Drug/Procedures Used:

Patients undergoing diagnostic catheterization at 10 sites in Massachusetts without on-site cardiac surgery and requiring nonemergent PCI were randomized in 3:1 fashion to either PCI at that institution or transfer to a site with on-site cardiac surgery (seven sites). To participate in MASS COMM, each hospital that did not have on-site coronary artery bypass grafting (CABG) was required to have approval from the Massachusetts Department of Public Health and meet minimum requirements for annual hospital and operator PCI volume.

Principal Findings:

A total of 3,691 patients were enrolled, 2,774 to PCI at sites without on-site cardiac surgery, and 917 to sites with on-site cardiac surgery. Baseline characteristics were similar between the two arms. Approximately 92% of the patients were white, 32% had diabetes mellitus, 22% had a prior history of myocardial infarction (MI), and 28% had a history of prior PCI. Of all patients, 64% were undergoing PCI for acute coronary syndrome (ACS). The majority of PCIs were in a single vessel (84%); three-vessel PCI was performed in 1.2% of patients only. Reference vessel diameter was slightly large in hospitals without on-site CABG (2.99 mm vs. 2.92 mm, p < 0.001). These hospitals were also more likely to implant bare-metal stents (32.6% vs. 24.6%, p < 0.001). Procedural success was similar (81.3% vs. 74.7%, p = 0.22).

The primary major adverse cardiac event (MACE) endpoint at 30 days was noninferior between sites without and with on-site cardiac surgery (9.5% vs. 9.4%, p for noninferiority < 0.001). Individual endpoints including all-cause mortality (0.7% vs. 0.3%, p = 0.44), MI (6.5% vs. 6.5%, p = 1.0), repeat coronary revascularization (2.7% vs. 3.5%, p = 0.25), emergent CABG (0.3% vs. 0.1%, p = 0.69), and stroke (0.4% vs. 0.1%, p = 0.21) were similar. At 12 months, rates of MACE were still noninferior (17.3% vs. 17.8%, p for noninferiority < 0.001). Once again, Individual endpoints including all-cause mortality (2.3% vs. 2.4%, p = 0.89), MI (8.6% vs. 7.8%, p = 0.55), repeat coronary revascularization (8.5% vs. 9.9%, p = 0.24), emergent CABG (0.3% vs. 0.1%, p = 0.69), and stroke (1.0% vs. 0.8%, p = 0.83) were similar. Other endpoints such as target lesion revascularization (4.9% vs. 5.0%) and stent thrombosis (1.1% vs. 2.1%) at 12 months were also similar.


The results of this trial indicate that nonemergent PCI at sites without on-site cardiac surgery has short-term (30 days) and intermediate-term (12 months) outcomes that are similar to those achieved at sites with on-site cardiac surgery. These results are identical to those noted in the larger CPORT-E trial. Current ACCF/AHA guidelines list PCI at sites without on-site cardiac surgery as a Class IIb indication, which may need to be reconsidered, as long as they adhere to the strict criteria listed in this trial (i.e., minimum procedural volume, quality improvement, participation in outcomes measurement projects).


Jacobs AK, Normand SL, Massaro JM, et al., on behalf of the MASS COMM Investigators. Nonemergency PCI at hospitals with or without on-site cardiac surgery. N Engl J Med 2013;Mar 11:[Epub ahead of print].

Presented by Dr. Alice Jacobs at ACC.13, San Francisco, March 11, 2013.

Clinical Topics: Acute Coronary Syndromes, Cardiac Surgery, Invasive Cardiovascular Angiography and Intervention, Aortic Surgery, Interventions and ACS

Keywords: Myocardial Infarction, Stroke, Acute Coronary Syndrome, Follow-Up Studies, Constriction, Pathologic, Percutaneous Coronary Intervention, Stents, Quality Improvement, Public Health, Thrombosis, Catheterization, Cardiac Surgical Procedures, Coronary Artery Bypass, Diabetes Mellitus

< Back to Listings