SUMMARY The ejection fraction is one of the most widely used measurements of left ventricular systolic function. Angiographic measurement of ejection fraction is based on determination of roentgenographic magnification and calculation of end-diastolic and end-systolic volumes, assuming a prolated ellipse. Because it is simple and radiographic magnification may not have been determined, some laboratories have obtained an "ejection fraction" by comparison of end-diastolic and end-systolic areas. A comparison of the two methods was made using ventriculograms of 538 patients from three cardiac catheterization laboratories. The area method of ejection fraction calculation consistently underestimates ejection fraction from left ventricular volumes. A regression equation was derived that allows adjustment of the ejection fraction obtained from areas to that from volumes.LEFT VENTRICULAR FUNCTION has been assessed by many methods. '-5 In the clinical setting, ejection fraction has been shown to be one of the most useful.6'-For example, it is related to prognosis as well as operative risk in patients with coronary artery disease and in those with other forms of heart disease.' [1][2][3][4][5] The angiographic technique for determination of left ventricular ejection fraction is based on the actual calculation of end-diastolic and end-systolic volumes, using the area-length method of Dodge and associates.'6' 17 These calculations require precise measurements of the distances between the left ventricle, the x-ray tube and image intensifier, and then filming a calibration grid at the level of the left ventricle.'8 Some laboratories omit this measurement of the roentgenographic magnification and report an "ejection fraction" based solely on comparison of enddiastolic and end-systolic silhouette areas. Studies have appeared without a description of the method for the determination of ejection fraction.'121 We compared the two techniques of calculating ejection fraction and present a regression curve that allows an accurate and simple determination of ejection fraction without having to take roentgenographic magnification and volume calculation into account. From MethodsThirty-degree right anterior oblique (RAO) left ventriculograms of 538 cases from three cardiac catheterization laboratories were reviewed. Patients with coronary artery disease, valvular disease, cardiomyopathies and congenital heart disease were included. The end-diastolic and end-systolic silhouettes were hand-traced. Volume MethodAssuming the ventricle to be ellipsoid, the volume of each silhouette was derived using the formulas of Dodge and associates.'6-8 Long-axis length (L) was measured from the mid-aortic root to the apex. Area (A) was measured by planimetry and minor axis (D) was calculated as 4A irLThe volumes (V) were then calculated by The areas of the end-diastolic and end-systolic silhouettes measured by planimetry were used to calculate an ejection fraction as

by Covvey and colleagues [I], we wish to comment on our experience with a computer-based catheterization laboratory system (CCLS). We agree that there is no commercially available CCLS that will meet all requirements of a given catheterization laboratory. However, we disagree with the assumption that one should proceed with the purchase of a CCLS only after all computer needs are recognized. We believe it is erroneous to assume that most individuals are able to recognize all their needs initially and would be able to measure the true impact of a computer system in the catheterization laboratory prior to the purchase.Three years ago we purchased a "turnkey" CCLS with on-line hemodynamic and off-line left ventriculogram analysis capabilities. Prior to this, we had had no experience with a CCLS. After reviewing the information available on each company's product, we realized that there was no one system available that could handle all of our needs. However, it was apparent that one system fulfilled most of our immediate requirements and was flexible enough to change and enhance as our sophistication grew.The CCLS was used for patient care within 48 hours of installation. Within a short time, through the use of parallel hand calculations, the reliability and accuracy of the system was documented. The time required for measuring and calculating hemodynamic data and performing left ventricular analyses has greatly decreased. Since the hemodynamic calculations are performed on-line, decisions concerning appropriateness and need for additional studies can be made while the patient is instrumented.Only after using a computer system in this environment for 6-8 months were we able to realize the true potential that this system could have in reducing physician and technician time. In addition, the system proved useful in standardizing and improvingpatient data used for therapeutic decisions. After two years we feel that we have gained enough experience to begin to enhance the operation of the system.The point we wish to emphasize is that with the use of a reasonable amount of logic and restraint, one can purchase a CCLS, utilize it immediately, and, after gaining experience, later enhance the software to meet specific needs. If the system to be purchased is flexible and can be utilized to perform most of the immediate applications, one does nor have to "be prepared to throw in the towel" or to spend needless man-hours developing a system just to be prepared for all potential applications.