No CrossRef data available.
Article contents
Interobserver variability of determination of the size of the left-to-right shunt in isolated atrial septal defects within the oval fossa by Doppler and cross-sectional echocardiography
Published online by Cambridge University Press: 19 August 2008
Summary
Although noninvasive estimation of the ratio of pulmonary to systemic flow by measuring the cross-sectional areas of the pulmonary and aortic valves and the respective time integrals of the velocity of blood is widely used to determine the amount of shunting in patients with left-to-right shunts, few data exist concerning interobserver variability. We assessed such variability of shunt calculations derived from Doppler echocardiography in 10 children aged six (3.5–11) years with an isolated atrial septal defect within the oval fossa. The ratio of flows was calculated by measuring stroke volumes over the pulmonary trunk and the aorta immediately before and at an average of seven days (5–11) following surgical closure of the atrial septal defect. Two independent observers measured the cross-sectional areas of the aortic and pulmonary orifices and the velocity time integrals and calculated the size of the shunt from these data. All measurements were performed five times, and the mean of these five measurements was calculated and compared for analysis. Before surgery, observer I had measured a ratio of pulmonary to systemic flow of 3.8 (2.1–6.2) and observer II of 3.5 (1.9–5). The mean interobserver difference in terms of the size of the shunt was 1.1 (or 30%). After surgery, observer I measured a ratio of 1.3 (0.8–1.7) and observer II of 1.1 (0.7–1.6). The cross-sectional area of the pulmonary valve was measured as 3.6 (2.6–4.9) cm2 and 3.4 (2.3–5.8) cm2 respectively, with a mean interobserver difference of 0.8 cm2 or 23%. For the velocity time integrals across the pulmonary trunk, the mean interobserver difference was 13% (4–38) before and 16% (0–55) after surgery. The respective variability for the aortic velocity time integrals was 11 % (4–46) before and 9% (0.5–23) after surgery. The variability of measurements of maximal flow velocity across the pulmonary trunk was 15% (4–38) before and 18% (1–77) after surgery. The respective variability for the aortic flow velocity was 11 % (2–28) before and 9% (4–23) after surgery. A paired t-test showed no significant differences for all parameters assessed. The difficulty of reproducing measurements of the pulmonary valve is an important factor affecting interobserver variability. The magnitude of interobserver difference was significantly related to the cross-sectional area of the pulmonary trunk (r=0.84; p<0.02). The interobserver variability may be considerable, and is large enough to influence clinical decisions. Thus, we conclude that a decision regarding need for surgery cannot be based solely on calculation of the shunt derived only from Doppler measurements in patients with isolated atrial septal defects within the oval fossa.
- Type
- Original Articles
- Information
- Copyright
- Copyright © Cambridge University Press 1992