Complications related to access site after percutaneous coronary interventions: are the adverse events underreported?

The primary objective was to compare the NCDR rate of four access site related complications (arteriovenous fistula, pseudoaneurysm, retroperitoneal bleed, and blood transfusion) with that obtained independently within a Quality Improvement (QI) Database. The secondary objective was to determine the in-hospital mortality related to these complications.

NCDR is the most comprehensive database of post-PCI patients and a method by which hospitals are compared for quality of care and outcomes. The collected data include in-hospital events only, therefore reporting for same day and/or early discharges may fail to capture adverse events. We hypothesized that the actual rate of complications post-PCI may be higher than the rate reported in NCDR.

Adverse events following PCIs were collected in a QI database. We compared the rate of four above mentioned complications between NCDR and our QI database for the time period between January 1, 2005 and December 31, 2008.

A total of 3,940 PCIs were performed on 3,430 patients in the four-year interval. The incidence of the combined endpoint of the four adverse events was 4.1% (161 events) in NCDR, vs. 4.7% (186 events) in QI database, a 13% higher yield. There was significantly higher in-hospital mortality in the complication group with regards to both all cause (4.7% vs. 1.1%, P < 0.0001) and cardiovascular mortality (2.7% vs. 1%, P = 0.046).

Access site related complications occurred 13% more than what was reported in NCDR and were associated with a greater than fourfold increase in in-hospital mortality.