Abstract 749: Moving 3D Echocardiography Into the Clinical Lab: How Often Are Standard Linear Dimensions and 3D Volumes Concordant for the Classification of Left Ventricular Size?
Background: Accurate classification of left ventricular (LV) size has important clinical implications for patients with a variety of cardiac conditions. When cardiac MRI is used as a gold standard, prior studies have demonstrated a strong correlation with volumes derived from 3D echocardiography. Since recent advances in echo technology have made a quantitative assessment of LV volumes in a busy clinical lab feasible, we investigated how often this technique results in a change in LV size classification as compared with traditional linear measurement methods in order to help gain insight as to one of the possible incremental values of 3D imaging in a real world setting.
Methods: We retrospectively identified 152 consecutive subjects who had 3D LV volumes obtained as part of a routine clinical echo (Age 59±16, 47% female, LVEF 54±14%, range 18–78%). All subjects had 3D quality deemed adequate by a staff echocardiographer such that LV volumes were included in the clinical report. The three most common indications were: CHF (20%), chemo related (23%), and miscellaneous (38%). LV size was classified according to the American Society of Echocardiography Guideline Document on Chamber Quantification using two methods: standard linear measurements (from m-Mode or 2D) and 3D volumes. The two methods were deemed “concordant” if they resulted in the same classification of LV size (i.e., mild, moderate, or severely dilated) and “discordant” if they did not. Normal EF was defined as ≥ 55%.
Results: Among study subjects with normal LVEF (N=88, 58%), there is a high concordance in LV size classification regardless of whether it is derived from linear dimensions or 3D volumes (LVIDd vs LVEDV = 90% and 92% when adjusted for BSA). However, among study subjects with an abnormal LVEF (N=64, 42%), there is low concordance in LV size classification when derived from linear methods vs those derived from 3D volumes (LVIDd vs LVEDV = 52% and 58% when adjusted for BSA).
Conclusion: Given that 3D echo has previously been well validated against cardiac MRI, our study suggests that in patients with abnormal LVEF, particular care should be taken to report LV size based on 3D volumes when possible as standard linear dimensions can lead to a misclassification in nearly half of these patients.