Complications occasionally occur as a result of primary PCI. Local vascular complications include bleeding, hematomas, pseudoaneurysms, and arteriovenous fistulae at the access site. These events occur in 2 to 3% of patients, about two thirds of whom require transfusion. Major bleeding (including bleeding at the access site) occurs in about 7% of patients undergoing the procedure. The incidence of bleeding has declined, probably because lower doses of heparin and smaller catheters are used now than in the past, as well as because of increasing experience among interventional cardiologists and ancillary personnel. The incidence of intracranial hemorrhage is lower with primary PCI than with fibrinolytic therapy (0.05% vs. 1%, P<0.001).
Severe nephropathy after PCI (caused, at least in part, by radiographic contrast material) occurs in up to 2% of patients. It occurs most often among those with cardiogenic shock or underlying renal insufficiency48 and those of advanced age. Anaphylactic reactions to radiographic contrast material are very rare.
Ventricular tachycardia or fibrillation is reported in 4.3% of patients undergoing primary PCI. Although these patients remain in the hospital longer than those who do not have ventricular tachyarrhythmias, the long-term prognosis for those with or without ventricular tachyarrhythmias is similar.
In patients undergoing elective balloon angioplasty, the abrupt closing of the infarct-related artery (during or within hours after the procedure) occurs in up to 3% of patients; it may occur even more frequently among those undergoing primary balloon angioplasty. Stenting of the infarct-related artery decreases the incidence of abrupt closing to about 1%, thereby diminishing the need for urgent bypass surgery and (in the opinion of some investigators) obviating the need for on-site surgical capability. Therefore, stenting is the preferred primary intervention if the coronary anatomy is suitable. As noted, stents also reduce the risk of restenosis, an effect shown to be even more marked with the use of drug-eluting stents. In most trials of stenting, stent thrombosis has occurred in less than 1.5% of patients receiving either a bare-metal stent or a drug-eluting stent within the first year.
Serious cardiovascular events occur in a small percentage of patients undergoing primary PCI. In the report of 4366 procedures described above, the rates of emergency cardiac surgery and in-hospital death were 4.3% and 2.5%, respectively. Such events occur much more frequently among patients in whom perfusion is not restored.
At centers where primary PCIs are performed, there is a direct relationship between procedural volume and outcomes. Among patients undergoing elective PCI at centers in which 200 or more such procedures are performed each year, the incidence of urgent bypass surgery and death is lower than among those whose procedure is performed at a center where fewer than 200 PCIs per year are performed.