Objective: To assess the impact of epiaortic scanning on the incidence of perioperative stroke in patients undergoing cardiac surgery.
Methods: Patients consecutively enrolled in our surgical database between January 2000 and August 2004 were subdivided into three groups depending on the planned use of epiaortic ultrasonographic scan. Patients treated before the availability of the equipment constituted group A (n=366). Epiaortic scanning was next performed selectively in group B (n=1116) and finally adopted on a regular basis in group C patients (n=690). Comparisons of stroke rates were performed in the whole series both according to the actual use of epiaortic scan and to the intended scan policy. A sub-analysis was additionally performed in the CABG cohort, where expected stroke rates could also be estimated by a validated model. Multivariable analysis was employed to identify predictors of early stroke.
Results: In the whole series, total stroke rates were 3.3%, 1.1%, and 1.9% for groups A, B, and C, respectively (p=0.02). Correspondingly, in the CABG cohort they were 3.4%, 0.5%, and 1.7%, respectively (p=0.002), with no substantial change following risk-adjustment. For the CABG cohort, total stroke rates were no different from expected estimates in the no-scan group A patients (3.4% vs 3.9%, ns). On the other hand, they were lower than expected in groups B+C (0.9% vs 2.8%, p=0.001), in patients actually scanned (1.4% vs 3.4%, p=0.01) and, among the latter, in those with significant aortic pathology (1.3% vs 4.5%, p=0.03). The risk reduction was particularly evident for early strokes, with no difference between scan groups: the rates were 0.5% and 0.6% for groups B and C, respectively, versus 2.2% in group A for the whole series (p<0.03), and 0.4% and 0.5% versus 1.9% in group A (p=0.02) for the CABG cohort. By multivariable analysis a no-scan policy (OR=4.0, 95% CL 1.4-11.4) and extracardiac arteriopathy (OR=3.0, 95% CL 1.1-8.0) were independently associated with early stroke.
Conclusions: The use of epiaortic scanning is associated with a lower risk of intraoperative adverse events leading to early postoperative stroke.