Echocardiography (ECHO) is useful to document changes in left ventricular mass (LVM) in groups of patients, but may be too variable for use in the individual patient. Magnetic resonance imaging (MRI) may be a more precise and reliable method to quantify the mass of the left ventricle. This study reports the accuracy, precision, and reliability of LVM estimates by MRI as compared to data obtained by ECHO in hypertensive patients.
Accuracy referred to the comparison of LVM by MRI to anatomical LVM determined by autopsy. Precision was examined using 34 duplicate MRI images and by blindly reading 24 duplicate M-mode strips. Reliability was assessed by MRI in four subjects over 2 months, and by ECHO in 22 hypertensive patients over 2 weeks. Agreement between MRI and ECHO estimates of LVM was determined in the same 17 hypertensive patients using linear regression.
MRI LVM estimates were within 17.5 g (95% CI) of the true LVM. The linear agreement between MRI and ECHO estimates of LVM could be described by the equation MRI = 0.61 × ECHO + 49.57 (r = 0.63, P < .01). The precision of LVM by MRI (11 g) was over twice that observed with ECHO (26 g). The reliability of MRI LVM estimates was more consistent (±8 g) than that for ECHO (±49 g).
MRI appears to be a more precise and reliable method for measuring LVM, and would be more suitable than ECHO for the clinical evaluation of the individual patient.
Am J Hypertens 1995;8:221–228