Abstract 1170: Incidence and Predictors of Stroke Associated with Percutaneous Coronary Intervention
The clinical features associated with stroke during percutaneous coronary intervention (PCI) have not been well defined. All patients undergoing PCI in NCDR from January 1, 2004 to March 30, 2007 were included (n=708,481) in this analysis. Stroke is defined in NCDR as central neurological deficit lasting >72 hours with onset anytime in the cardiac catheterization laboratory until hospital discharge. A multivariate model was constructed that included medical history, risk factors for coronary disease, and angiographic variables. Adjusted odds ratios for development of stroke with PCI were calculated by generalized estimating equation. Stroke developed in 0.34% patients (n=2,438). Patients who developed a stroke were more likely to be older, female, and have greater prevalence of acute coronary syndromes and concomitant medical illnesses including old MI, heart failure, and known cerebrovascular disease, compared with patients who did not develop a stroke. Patients with stroke had worse left ventricular systolic function, greater incidence of high risk lesions and worse TIMI 3 flow rates. In multivariate analysis, admission with ST elevation MI (Odds ratio, OR=2.61; 95% CI = 1.33–5.14) and use of intraaortic balloon pump (OR=2.67) were strongest independent predictors for development of stroke in association with PCI. In-hospital mortality was greater in patients who developed a stroke compared with those without a stroke (14% versus 1%, p<0.0001), and so were peri-procedural myocardial infarction and bleeding outcomes. Stroke developing in association with PCI is rare but carries high in-hospital mortality. Clinical characteristics can be used to predict individuals at the highest risk of developing this complication. The association of stroke following PCI with older age, female gender, and extensive coronary disease is not unexpected, but the strong association with acute coronary syndromes is noteworthy.