Pii: S1010-7940(01)00931-9

Background: The Multicenter Study of Perioperative Ischemia (McSPI) developed and validated a Stroke Risk Index for estimating the likelihood that patients undergoing isolated coronary artery bypass grafting will experience major perioperative neurologic events. The International Council of Emboli Management (ICEM) Study Group has suggested that use of intraaortic filtration reduces adverse neurologic events. Objective: The objective of the present study was to compare predicted and observed neurologic outcomes in patients receiving intraaortic filtration. Patients and Methods: From February 1999 to August 2000, 962 patients were enrolled consecutively in a prospective, nonvoluntary registry of intraaortic filtration in 15 European centers. Of these, 447 underwent isolated coronary artery bypass grafting, the target population for applying the McSPI Stroke Risk Index. Forty-five had incomplete data, yielding a study group of 402 patients. The Stroke Risk Index was calculated for each patient, and the sum across patients yielded an expected number of neurologic events. These were compared to observed events by confidence limits and goodness of fit. Results: Six neurologic events were observed (1.5%; 95% confidence limits 0.6–3.4%), roughly half the 13.7 predicted by the Stroke Risk Index (3.4%; 95% confidence limits 2.0–5.8%), P 1⁄4 0:03. Conclusions: Adverse neurologic events associated with coronary artery bypass grafting in which intraaortic filtration was used were rare and fewer than expected on the basis of the Stroke Risk Index. Rare occurrence of clinically relevant events precludes their use as primary endpoints for randomized clinical studies; however, the Stroke Risk Index provides a valuable benchmark in the absence of such studies. q 2001 Elsevier Science B.V. All rights reserved.