Intraoperative Myocardial Ischemia

Based primarily on results obtained during exercise treadmill testing, electrocardiographic (ECG) leads II and V5 are the suggested optimal leads for detecting intraoperative myocardial ischemia. However, these recommendations have not been validated in this setting using all 12 ECG leads. Accordingly, the authors studied 105 patients with known or suspected coronary artery disease (CAD) undergoing noncardiac surgery with general anesthesia by continuously recording the 12-lead ECG intraoperatively in all patients. The average duration of monitoring was 8.2 +/- 2.7 h (mean +/- SD). Ischemic episodes (i.e., greater than or equal to 1-mm horizontal or downsloping ST depression, greater than or equal to 1.5-mm slowly upsloping ST depression or greater than or equal to 1.5-mm ST elevation in a non-Q wave lead) occurred in 25 patients (24%). Out of 51 ischemic episodes, 45 involved ST depression alone, and the remaining six involved both ST depression and elevation. ST segment changes occurred in a single lead only in 14 episodes, while multiple leads were involved in 37 episodes. Lead sensitivity was estimated assuming that all ST segment changes were true positive responses. Sensitivity using a single lead was greatest in V5 (75%) and V4 (61%), and intermediate in II, V3, and V6 (33%, 24%, and 37%, respectively). The remaining seven leads demonstrated very low sensitivity (2-14%) or exhibited no ischemic changes (I and a VL). Combining leads V4 and V5 increased sensitivity to 90%, while the standard clinical combination, II and V5, was only 80% sensitive. Sensitivity increased to 96% by combining II, V4, and V5. The further addition of V2 and V3 (five leads) increased sensitivity to 100%. This study confirms previous recommendations for the routine use of a V5 lead (either uni- or bipolar) in all patients at risk for ischemia. V4 is more sensitive than lead II, and should be considered as a second choice. However, lead II, superior for detection of atrial dysrhythmias, is more easily obtained with conventional monitors. The use of all three would appear to be the optimal arrangement for most clinical needs, and is recommended if the clinician has the capability.