abstract & commentary
Synopsis: Interstudy variability was substantial for the color Doppler assessment of the severity of AR and MR and were related to changes in blood pressure and regression to the mean.
Source: Gottdiener JS, et al. Am Heart J. 2002;144: 115-121.
Color doppler echocardiography is the principle method for the serial evaluation of patients with left heart valve regurgitation, yet substantial variability in the estimated severity of regurgitation has been demonstrated. Thus, Gottdiener and associates sought to determine the relative contribution of acquisition vs. reader variability in the serial assessment of aortic regurgitation (AR) and mitral valve regurgitation (MR). The 23 subjects enrolled in the study ranged in age from 18 to 65 years and all had a body mass index > 27 kg/m2 since they were derived from the control group of an obesity pharmacologic study. The subjects were selected for an even distribution of aortic and mitral regurgitation from none to severe. Using the same machine, the same sonographer performed 2 echoes 14 days apart and then a different sonographer repeated the second echo within 2 hours (third echo). Mitral regurgitation was graded as none, trace, mild, moderate, or severe using the color flow jet area to left atrial area ratio method. Aortic regurgitation was graded using the ratio of height of the regurgitant jet to left ventricular outflow tract area. All echoes were read by 2 blinded experienced readers in a random sequence and 9 were read twice by the same reader. Acquisition variability was determined as the total variability minus the intrareader variability. Blood pressure at the time of the echo was measured and showed considerable variability between studies. Total variability was 29% for AR and 25% for MR. Intrareader variability was 6% for AR and 17% for MR. Acquisition variability was 29% for AR and 8% for MR. Predictors of total variability included initial regurgitation grade; the higher the initial severity the more likely severity would decrease on the second exam and vice versa. Also, diastolic blood pressure was directly related to changes in severity of both MR and AR. Gottdiener et al concluded that interstudy variability was substantial for the color Doppler assessment of the severity of AR and MR and were related to changes in blood pressure and regression to the mean.
Comment by Michael H. Crawford, MD
Changes in valvular regurgitation severity are important for assessing the response to specific therapy such as vasodilators, determining the timing of corrective surgery and evaluating changes in patient symptoms. Despite the importance of such decisions, the common use of color Doppler echo techniques to assess regurgitation severity are known to be problematic. The cardiology literature is filled with studies of new ways to assess regurgitation severity, each more complicated and problematic than the next. This has prompted some leaders in the field to recommend multi-measurement assessment systems, but the time required to do all the required measurements in the usual clinical practice makes this approach impractical. So most labs use the techniques studied in this report; jet area for MR and jet height for AR.
Considering that this was a research study performed under carefully controlled circumstances, the variability observed was considerable; 29% for AR and 25% for MR. The components of this variability was somewhat different for the 2 lesions. Echo acquisition issues were more prevalent with AR, including biological variability, mainly in blood pressure. Whereas, with MR, intrareader variability was more of a problem with considerable regression to the mean observed. The latter phenomena have been well studied and validated. Basically if a measurement is erroneous, a repeat measurement is unlikely to be more erroneous, so there is a shift toward the mean value. In the case of valve regurgitation severity this could be interpreted as a biologic change. Although not studied here, interreader variability is likely to be greater than intrareader. Thus, if this study was done under the usual clinical situation of different machines, sonographers, and readers, the results would probably have been worse. Clearly we need to do a better job reading serial studies in patients with valvular regurgitation.
This study has several limitations. The number of subjects is small. The technicians knew this study was focusing on MR and AR. Only 1 method of assessing regurgitation severity was used for each valve. Also, most of the subjects had trace-to-mild regurgitation. Several conclusions can be drawn, however, as pointed out by Dr. Schiller in his editorial. First, blood pressure should be measured and considered in the interpretation of the echo. Second, multiple methods of severity estimation should be used whenever possible. Third, changes in the trace-to-mild range should be downplayed. In my experience there is little difference between 1-2+ regurgitation (scale of 4 being severe). Finally, clinical decisions should take into account whether there are corresponding changes in LV and LA volume, LV function, estimates of filling pressure, and pulmonary artery pressure. Changes in 1 echo Doppler parameter alone are usually not sufficient to make major therapeutic changes.
Dr. Crawford is Professor of Medicine, Mayo Medical School; Consultant in Cardiovascular Diseases, and Director of Research, Mayo Clinic, Scottsdale, AZ.