Coronary angiograms and myocardial scintigraphy

Abstract

To the Editor.— We read the work by Galbraith et al (240:2053, 1978) concerning interobserver variability in the interpretation of coronary angiograms. These authors correlated the findings on premortem coronary angiograms of 20 patients with the postmortem pathological findings in fresh, unfixed heart specimens from the same patients. The authors used as a criterion for "substantial" lesions a reduction in lumen diameter of a given vessel of 50% or greater angiographically, or 75% reduction in cross-sectional area pathologically. Their basic results indicate that while minimal interobserver variability was present, there was considerable overall inaccuracy in the interpretation of coronary angiograms, with false-positive misinterpretations being the most common (about twice the false-negative rate). Within the context of their study, in which they set forth specific criteria for correlation of angiographic and pathological findings, the results are straightforward and of utmost validity. However, we wish to point out that in the clinical