Observer variation in the angiocardiographic diagnosis of mitral valve prolapse

The purpose of this study was to assess the reliability of the angiocardiographic diagnosis of mitral valve prolapse by measuring agreement between observers using defined diagnostic criteria. Sixty high-quality left ventriculograms, selected to include many that showed possible mitral valve prolapse, were reviewed by three angiocardiographers. Disagreement between observers as to positivity occurred in 26 percent and 30 percent of the cases on two reviews, and disagreement as to specific scallop involvement occurred in 68 percent and 78 percent of the cases. A specific observer agreed with himself in interpretation of positive 78 to 82 percent of the time and in the interpretation of specific scallop involvement 55 to 90 percent of the time. These results indicate that in the absence of generally agreed upon quantitative angiocardiographic criteria for the diagnosis of mitral valve prolapse, there is considerable variability among observers in the interpretation of mitral valve prolapse by angiocardiographic studies.