From ACC.13: NCDR Registries Reveal Excesses and Deficits in Cardiac Care Patterns

The ACC's National Cardiovascular Data Registry (NCDR®) collects clinical data in order to improve CV care, helping health systems, individual hospitals, and even clinical practices answer important questions about the care they deliver. Among this year's NCDR sessions at the 62nd Annual Scientific Session & Expo of the ACC were key reports revealing critical gaps in cardiac care, which are important to know given impending changes that will place new emphasis on quality of care. In our special coverage, Krishna Aragam, MD, discusses low rates of referral for cardiac rehabilitation; Vivian Tsai, MD, shows low rates of ICD-R implantation in eligible older patients; Jonathan C. Hsu, MD, details unwarranted exposure of patients to bleeding risk through inappropriate anticoagulation in low-risk patients; and Robert N. Vincent, MD, reviews adverse event patterns in pediatric and adult patients undergoing diagnostic catheterizations for congenital cardiac defects.

From The CathPCI Registry®

Post-PCI, Referral to Cardiac Rehabilitation May Be Too Often Overlooked
Cardiac rehabilitation (CR) aims at improving comprehensive CV risk reduction through supervised exercising training, patient education, nutritional and psychological counseling, and overall lifestyle modification. Benefits of CR in functional status and event reduction have been amply demonstrated in studies. In a retrospective analysis of prospectively collected data on 2,395 consecutive PCI patients, Kashish Goel, MD, and colleagues identified a 47% reduction in all-cause mortality in patients undergoing CR.1

Starting in 2014, The Centers for Medicare and Medicaid Services (CMS) will require reporting on referral to CR as a quality-of-care measure for patients with qualifying events (i.e., AMI or PCI). Moreover, CR referral is a newly proposed ACCF/AHA PCI performance measure.

Dr. Aragam's analysis assessed CR referral trends, patterns, and predictors in a study population of 1,432,399 patients from the CathPCI Registry® undergoing PCI between 2009 and 2011 in 1,310 US hospitals. They found:

  • Over the analysis period, the cumulative referral rate for all patients was 59.2%; for AMI patients with Medicare it was 66.0%. "Clinically, it's a very minor difference," Dr. Aragam noted.
  • Multivariate analysis showed that prior PCI, diabetes, peripheral artery disease, older age, prior CABG, and lack of insurance all predict lower referral rates. Acute presentation and private insurance increased odds of referral to CR.
  • Among quality-of-care measures, CR referral was substantially lower (~60%) than other such measures (~80–98%), including ACE/ARB among patients with an EF <40%, aspirin, beta-blocker, statin, and any P2Y12 medications at discharge.
  • Interhospital variation in referral rates was substantial—with most hospitals having CR rates of either >80% or <20%. Interestingly, insurance status had little bearing on CR referral rates.

"I think this study provides a good baseline of where we are before these measures are instituted. That will allow us to see how we do over time in comparison," Dr. Aragam said.

From the ICD Registry™

Cardiac Resynchronization Therapy in the Elderly
Most subjects studied in clinical trials of cardiac resynchronization therapy and defibrillator (CRT-D) are younger than those typically seen in "real-world" practices, where mean patient age continue to rise. Using NCDR ICD Registry™ data from 2006 to 2009 to identify patients who met clinical trial inclusion criteria for CRT implantation, Dr. Tsai and colleagues aimed to describe CRT-D use and its impact on survival in patients ≥65 years of age. Among their findings:
  • Study patients had QRS duration >120 msec and LVEF ≤ 35% (NYHA class III or IV).
  • The analysis included about 22,800 patients 65–74 years of age, an additional 22,700 patients 75–84 years, and about 4,600 patients >85 years.
  • The CRT/ICD split was about 83%/17% in both the 65–74 and 75–84 age groups, with a roughly 80%/20% split for CRT-D/ICD in the oldest group.
  • More than 80% of patients ≥65 years who received ICDs were also CRT candidates.
  • Mortality was higher among ICD recipients versus CRT-D patients.

"Everyone in this group met inclusion criteria for receiving CRT-D devices, and it's unclear why they didn't receive them," Dr. Tsai said in an interview. In some cases, inability to place the necessary third lead might have been a factor, she speculated. "Essentially, patients who received CRT-D had significantly better survival, both under and over 65 years of age."

From the IMPACT Registry®

Improving Pediatric and Adult Congenital Treatment
The newest NCDR database is the IMPACT Registry®, which gathers voluntary information on all cardiac catheterizations in pediatric and adult congenital heart disease patients. Investigators analyzed patient and procedural information from January 2011 through September 2012 on diagnostic and interventional cardiac catheterizations; this included 12,037 cath lab visits (32.8% diagnostic; 67.2% interventional) at 64 sites. While diagnostic catheterization is associated with low rates of adverse events in adults, that is not the case in very young children. Dr. Vincent and colleagues also found:
  • Procedures were performed on 292 newborns (<30 days), 973 infants (between the ages of 30 days and 1 year), 1,873 children (>1 year but ≤18 years), and 708 adults (≥18 years).
  • Procedures were classified as elective (86%), urgent (12%), emergent (2%), or salvage (<1%). General anesthesia was used in 76.5% of cases.
  • In diagnostic procedures, adverse events through discharge or 30 days were reported in 9.1% of patients. Adverse events were most common in newborns (28.8%) and infants (15.5%), and less common in children (4.8%) and adults (4.4%).
  • Also in diagnostic catheterizations, all-cause hospital mortality was 3.5%. (See Figure for breakdown by age.)
  • Major adverse events (cardiac arrest, tamponade, air embolus, embolic stroke, pacemaker, unplanned cardiac surgery, and death) were reported in 4.4% of patients, with 21.0% occurring in newborns, 6.5% in infants, 1.3% in children, and 2.1% in adults. Dr. Vincent said, "Although adverse events as currently reported are not necessarily attributable to the catheterization procedure itself, it is clear that newborns and infants who require a diagnostic catheterization are at risk of adverse events, including death, during their hospital admission." He added that future research will address risk stratification, attribution of adverse events and long-term follow-up of interventions, and creation of recommendations to guide catheter intervention.

From the PINNACLE Registry®

Unnecessary Oral Anticoagulant Use in Low-Risk AF Patients
"There's been a lot of focus on individuals with high risk of thromboembolism not getting appropriate anticoagulation," said Dr. Hsu in a presentation at ACC.13, but there are AF patients—"some call them lone AF patients"—with no thromboembolic risk who are being anticoagulated despite consensus guidelines that caution against it. "We wanted to see what's happening in the United States on a day-to-day basis with those individuals."

Dr. Hsu and colleagues examined records of patients enrolled in the PINNACLE Registry® between July 2008 and June 2012 who met criteria (age <60 years, CHADS2 score = 0, and no structural heart disease) for an ACC/AHA/ESC class III indication against use of anticoagulation. Their aim was to determine the prevalence and predictors of oral anticoagulation use in these low thromboembolic risk AF patients. Among their findings:
  • Among 7,661 patients identified as low-risk for thromboembolism, anticoagulation was prescribed in 1,694 or 22.1% (warfarin = 89.4%; rivaroxaban or dabigatran = 10.6%).
  • Multivariate analysis showed older age, male gender, and Medicare insurance (versus private insurance) predicted oral anticoagulation.
  • Also, multivariable analysis showed treatment in Southern states or the Western United States was associated with less frequent inappropriate use of oral anticoagulation compared to Northeastern states (RR = 0.61 and RR = 0.58, respectively).

Dr. Hsu concluded, "In a large, real-world population of low thrombotic risk outpatients with AF, one in five were treated with oral anticoagulation against guideline recommendations."

In an interview, he described a typical patient receiving inappropriate anticoagulation as being a young male with lone AF and normal LVEF on echocardiographic exam with an active lifestyle and no other risk factors or comorbidities. "We are unnecessarily submitting many patients to risks of major and life-threatening bleeding and death without any antithrombotic benefit. Especially with warfarin, intracranial hemorrhage is our biggest fear—and these people are at risk for that," Dr. Hsu said.

The next step, he continued, is to figure out the provider-level reasons for unnecessary anticoagulation.


1. Goel K, Lennon RJ, Tilbury RT, et al. Circulation. 2011;123:2344-52.

Keywords: Follow-Up Studies, Morpholines, Referral and Consultation, Peripheral Arterial Disease, Heart Arrest, Cardiac Resynchronization Therapy, Thromboembolism, Registries, Prevalence, beta-Alanine, Outpatients, Benzimidazoles, Stroke, Hospital Mortality, Multivariate Analysis, Risk Reduction Behavior, Cardiac Catheterization, Centers for Medicare and Medicaid Services, U.S., Intracranial Hemorrhages, Defibrillators, Implantable, Diabetes Mellitus

< Back to Listings