Michael O'Riordan
December 03, 2014

DURHAM, NC — An analysis of nearly two million PCIs shows the rate of bleeding varies extensively across US hospitals, ranging from 2.1% at the best-performing hospitals to just over 10% for the worst[1]. The variation was attenuated after adjustment for patient characteristics, but substantial differences in post-PCI bleeding rates still remained across US centers, say investigators.

The use of bleeding-avoidance strategies, such as using bivalirudin, performing PCI via the radial artery, or using vascular closure devices, also varied among the 1292 hospitals included in the CathPCI Registry.

The analysis, which was led by Dr Connie Hess (Duke Clinical Research Institute, Durham, NC) and published November 25, 2014 in Circulation: Cardiovascular and Quality Outcomes, supports the use of PCI-related bleeding complications as a performance measure in the CathPCI Registry and the potential incorporation of the quality metric into other PCI registries, say investigators.

Overall, the unadjusted median rate of bleeding was 5.2%. For centers in the fifth and 95th percentile, the bleeding rates ranged from 2.6% to 10.4%, respectively. When the CathPCI Registry risk model was applied to adjust for case mix, the bleeding rates in the fifth and 95th percentile were 2.8% and 9.4%, respectively. Median bleeding rates at academic and nonacademic hospitals were similar (5.4% vs 5.1%, respectively). Increasing use of heparin or GP IIb/IIIa inhibitors were associated with greater bleeding risk, while use of bivalirudin, radial PCI, or a vascular-closure device reduced the bleeding risk.

"Our results also suggest that provider decisions on procedure methods, such as [bleeding-avoidance strategies],may be useful to reduce PCI bleeding," according to Hess and colleagues. "Ultimately, quality-improvement initiatives to reduce post-PCI bleeding, perhaps through wide implementation of [bleeding-avoidance strategies] and sharing of practices from best-performing sites, might lead to improved PCI outcomes, although further investigation is needed."

Source: http://www.medscape.com/viewarticle/...80112PN&spon=2