Reproducibility of centralized laboratory interpretations of measures of myocardial infarct size, ventricular function, and ventricular electrical instability scintigraphy

Abstract

In many clinical studies of acute myocardial infarction, measurements such as infarct occurrence and size and ventricular function have proven useful in the assessment of the effects of new therapies. The importance of these measures is increasing as mortality from acute infarction continues to decrease. In the Multicenter Investigation of the Limitation of Infarct Size study of 985 patients with suspected or confirmed acute myocardial infarction, several such measures were obtained in centralized core laboratories to reduce interlaboratory variations. In particular, core laboratories were responsible for measurement of serial changes in plasma creatine kinase activity, pyrophosphate scintigrams, ambulatory ECGs, radionuclide ventriculograms, and ECGs. To quantitate the variability in measurements made by these core laboratories, duplicate, blinded data were submitted to the laboratories on 88 patients. In general, the five core laboratories had reliability coefficients of about 90%. The analysis of these quality control data supports the feasibility of monitoring core laboratories with duplicate samples. They also provided quantitative information on the variability of the duplicate measurements and on the relative magnitude of patient-to-patient variability versus measurement error. Because measurement errors between duplicate readings generally were small, these data provide additional verification of the data from the multiple publications of the Multicenter Investigation of the Limitation of Infarct Size study.