Background: ST monitoring with continuous 12-lead ECG is a well-established method in patients with unstable coronary artery disease (CAD). However, the method lacks documentation on optimal criteria for episodes of transient ischemia and on observer variability.
Methods: Observer variability was evaluated in 24-hour recordings from 100 patients with unstable CAD with monitoring in the coronary care unit. Influence on ST changes by variations in body position were evaluated by monitoring 50 patients in different body positions. Different criteria of transient ischemia and their predictive importance were evaluated in 630 patients with unstable CAD who underwent 12 hours of monitoring and thereafter were followed for 1 to13 months. Two sets of criteria were tested: (1) ST deviation > or = 0.1 mV for at least 1 minute, and (2) ST depression > or = 0.05 mV or elevation > or = 0.1 mV for at least 1 minute.
Results: When the first set of criteria were used, the interobserver agreement was good (kappa = 0.72) and 8 (16%) had significant ST changes in at least one body position. Out of 100 patients with symptoms suggestive of unstable CAD and such ischemia, 24 (24%) had a cardiac event during follow-up. When the second set of criteria were used, the interobserver agreement was poor (kappa = 0.32) and 21 (42%) had significant ST changes in at least one body position. Patients fulfilling the second but not the first set of criteria did not have a higher risk of cardiac event than those without transient ischemia (5.3 vs 4.3%).
Conclusions: During 12-lead ECG monitoring, transient ischemic episodes should be defined as ST deviations > or = 0.1 mV for at least 1 minute, based on a low observer variability, minor problems with postural ST changes and an important predictive value.