Study: CABG use decreased between 2001 and 2008

Epstein A. JAMA. 2011;305:1769-1776.

Issue: June 2011

Results from an analysis published in JAMA suggested that CABG use decreased between 2001 and 2008, but also that percutaneous coronary intervention remained at stable levels.

Researchers for the cross-sectional study examined time trends in patients undergoing CABG or PCI between 2001 (n=77,952 for CABG; n=55,483 for PCI) and 2008 (n=151,893 for CABG; n=153,800 for PCI) in the Healthcare Cost and Utilization Project’s Nationwide Inpatient Sample. The data were supplemented by Medicare outpatient hospital claims.

According to the results, there was a 15% decrease in coronary revascularizations, from 5,569 (95% CI, 5,315-5,835) procedures/million-adults in 2001-2002 to 4,748 (95% CI, 4,532-4,975) procedures/million-adults in 2007-2008 (P<.001). Annual rates of annual CABG surgery decreased by 38%, from 1,742 (95% CI, 1,663-1,825) procedures/million-adults in 2001 to 1,081 (95% CI, 1,032-1,133) procedures/million-adults in 2008 (P<.001). The decreases were observed across sex, age racial and regional subgroups.

By contrast, annual rates of PCI decreased by 4%, from 3,827 (95% CI, 3,578-4,092) procedures/million-adults in 2001-2002 to 3,667 (95% CI, 3,429-3,922) procedures/million-adults in 2007-2008 (P=.74).

“Although the total rate of US coronary revascularization decreased modestly between 2001 and 2008, there was a substantial decrease in the CABG surgery rate,” the researchers concluded. “Between 2001 and 2008, the rate of PCI did not significantly change; however, there were continual changes in the frequency of stent types used for PCI.”

Riley R.Circ Cardiovasc Qual Outcomes. 2011;4:193-197.

As we move forward into an era of limited health care resources, it will become increasingly important to have an accurate estimation of the clinical need and utilization of medical procedures. This information will enable forecasting of needs for resource allocation and decision-making about future investments in clinical infrastructure and training of future physicians.

In the study published by JAMA, Epstein et al estimate that rates of CABG have declined steadily by 15%, while PCI rates have declined only slightly, with a non-significant 4% decrease in 2008 compared with 2001. Although they claim that PCI volumes remained steady, their data show PCI procedures increased from 2001 to 2004, but decreased by 10% from 2005 to 2008.

Our similar study published in Circulation: Cardiovascular Quality and Outcomes also found a steady decrease in CABG, but a more complicated trend in PCI procedures. We reported annual CABG and PCI revascularization rates in the entire Medicare population from the beginning of 2000 through the end of 2009 and found an early increase in PCI procedures from 2000 to the end of 2004, but a steady decrease in PCI procedures from 2004 through 2009. Epstein et al also suggest a rise in multivessel stenting. Our data show that any increase multivessel stenting peaked in 2006, before publication of the SYNTAX study, and has since declined. Based on the rate of decrease in PCI, we concluded the decrease in CABG volume could not be entirely accounted for by a preferential use of PCI.

Epstein et al’s conclusion that PCI rates have remained unchanged must be evaluated closely. Both their study and ours show the changes in PCI volume were not steady over the 2000 to 2008 time period. However, their comparison between the 2001 and 2008 data assumes a linear change and, therefore, does not accurately reflect the trends that occurred in the intervening years.

The minor differences in the procedure rates published in our study result from use of different billing datasets and methodology. We utilized nearly complete Medicare physician billing data on inpatient and outpatient procedures, whereas Epstein et al evaluated all insurance types, but did not have data on outpatient procedures and made estimations of rates for their sample based on the US population, rather than the number of patients actually eligible for treatment in this subset of hospitals.

Importantly, Epstein et al acknowledge that the present study does not provide sufficient patient-level data to form any conclusions regarding procedural appropriateness. These revascularization rates may reflect changing patient demographics and clinical comorbidity within this sample of hospitals, rather than operator or patient preferences. An increase in acute coronary syndromes in numbers of patients with inoperable/high-risk surgical disease or low-risk non-surgical disease or the migration of patients from one hospital to another may explain their findings.

– Creighton W. Don, MD, PhD

Assistant Professor of Medicine, Division of Cardiology

University of Washington, Seattle

Member, Society for Cardiovascular Angiography and Interventions

Disclosure: Dr. Don reports no relevant financial disclosures.

Follow CardiologyToday.com on Twitter.