Inter- and intra-observer variability in detection of patent foramen ovale with contrast echocardiography.

Abstract

Patent foramen ovale is detected in-vivo using the technique of contrast echocardiography. A positive test depends on the ability of an observer to recognise micro-bubbles passing from the right to the left heart, and to differentiate them from background noise and/or chordae tendinae. To assess the inter- and intra-observer variability of this technique four observers reviewed independently, and on two separate occasions, 100 consecutive contrast studies stored on video tape. The proportions of agreement (95% CI) for each separate observer on the two occasions for the presence of patent foramen ovale were 0.91 (0.76-0.98), 0.88 (0.73-0.92), 0.80 (0.63-0.92) and 0.78 (0.60-0.91), and for its absence were 0.96 (0.88-0.99), 0.95 (0.87-0.99), 0.90 (0.81-0.96) and 0.91 (0.82-0.96), respectively. The proportions of agreement (95% CI) between all observers on the first viewing were 0.78 (0.72-0.84) for the presence and 0.91 (0.87-0.93) for the absence of patent foramen ovale. For the second viewing these were 0.81 (0.75-0.86) and 0.91 (0.88-0.94), respectively (P = NS, Chi-squared test). These results indicate that contrast echocardiography is subject to clinically acceptable inter- and intra-observer variability.