Failed percutaneous coronary intervention: A decade of experience in 21,000 patients

Objectives: The purpose of this study was to evaluate the incidence, mechanisms, and in‐hospital outcomes after failed percutaneous coronary intervention (PCI). Background: The most common causes of PCI failure are severe dissection, no‐reflow, and uncrossable total occlusions. Methods: Demographic, clinical, procedural, and quality assurance data were recorded in a prospective database to determine the incidence and outcome of failed PCI. Data were analyzed according to the year of PCI: prestent era (PSE 1993–1995), first generation stent era (FGSE 1996–2000), and contemporary stent era (CSE 2000–2003). Results: PCI (20,884) were performed over a 10‐year‐period from 1993 through 2003, including 4,344 (20.8%) in the PSE, 9,114 (43.6%) in the FGSE, and 7,426 (35.6%) in the CSE; 19,171 (91.8%) were successful and 1,704 (8.2%) were failures. Nearly 90% of PCI failures are due to no‐reflow or uncrossable total occlusions. Among PCI failures, 7.5% were treated medically and 0.7% required emergency coronary artery bypass grafting (ECABG). PCI success was independently associated with PCI during CSE (OR 1.86, P < 0.0001). ECABG decreased sixfold from 1.2% in the PSE to 0.2% in the CSE. Conclusions: Contemporary PCI patients have better procedural outcomes and fewer in‐hospital adverse events than patient treated before the availability of stents. However, recanalization of total occlusions and avoiding no‐reflow will have the most impact on future PCI success. © 2008 Wiley‐Liss, Inc.