Comparison of visual estimate with digital caliper measurement of coronary artery stenosis

We compared visual estimate and digital caliper measurement of coronary stenoses, utilizing both a 50% and 70% diameter reduction as a definition of significance and both experienced angiographers and cardiology fellows as readers. Ten angiograms were interpreted twice, using the different methods, by four readers, two months apart. The angiograms were divided into 12 vessel segments. Stenoses were judged more severe by visual estimate in 11 segments. Seven of twelve segments had significant differences between readers using the visual estimate, but no significant differences were obtained using caliper measurements. Variability was highest for fellows using the visual method. Interobserver agreement was highest using the digital caliper and the 70% criteria. Intraobserver agreement was most dependent on the amount of time taken in film interpretation. The digital caliper appears to be preferable in interpreting angiograms, particularly for cardiology fellows.