First-third ejection fraction: is the first-pass radionuclide method accurate?

Previous reports have suggested that left ventricular first-third ejection fraction (EF) can be obtained from the left ventricular time-activity curve derived from first-pass radionuclide angiography based on Anger camera data. The validity of this technique was assessed by: a study of beat-to-beat variations in data from 15 patients in which electrocardiographic data were simultaneously recorded, and a computer simulation incorporating the application of Poisson statistics to appropriate count rate data. The results of patients studies showed no consistent trend in any first-third parameter obtained from consecutive beats in individual subjects, and unacceptably high statistical uncertainty in the calculation of the first-third ejection fraction. The weighted standard deviation of the first-third ejection fraction in each of 15 patients studied averaged 7.5 EF units, while first-third ejection fraction averaged 22.9 EF units. The relative error averaged 32%. The computer simulation indicated a high relative error of 47% associated with the first-third ejection fraction at typical end-diastolic count rates of 200 per frame from 1,000 computer Poisson randomizations of an appropriate analog volume curve. The results render the first-pass radiocardiographic method invalid for determining first-third ejection fraction.