Background: We assessed the accuracy of two electron beam computed tomography (EBCT) protocols for predicting coronary events.
Methods and results: In 1994, 24 months after enrollment in a longitudinal study, 326 high-risk adults underwent both 3- and 6-mm image-slice thickness EBCT scanning and were followed up for 32.0+/-4.0 additional months. Events were defined as either coronary death, myocardial infarction, or revascularization. We monitored these subjects for the 32-month postscanning period with yearly phone calls and acquisition of records for all hospital admissions. At the time of scanning, 11 subjects (3%) had already suffered 12 events (5 infarctions and 7 revascularizations) during the 24-month prescanning period. During the postscanning period, 18 subjects (6%) suffered 23 events (5 coronary deaths, 6 infarctions, and 12 revascularizations). Thus, 28 subjects (9%) suffered 35 events. Calcium quantities calculated for both protocols, performed on the same subjects, were sorted in ascending order and divided into equal quartiles. When revascularizations were included, there was a significant trend toward higher frequencies of events with increasing calcium quantity (P<.01). However, coronary death and infarction were not significantly more frequent in higher quartiles. These relationships were preserved in the subjects without prior events at the time of scanning.
Conclusions: Calcium quantities from the 3-mm and the more reproducible 6-mm scanning are equally accurate for predicting events. Coronary calcium amount appears to be a weak predictor of coronary death and infarction. Its predictive accuracy is superior for predicting revascularization.