The reproducibility of intrapartum cardiotocogram assessments

Five obstetrician-gynecologists experienced in fetal monitoring assessed 150 intrapartum cardiotocograms obtained with an external transducer. There were three successive blind readings, the first two without any clinical data apart from gestational age. The reviewers indicated whether the tracings showed definite, possible or no abnormalities. The interpretations given by each reviewer for any one tracing were fairly consistent, but they varied markedly from one reviewer to another. The proportion of tracings interpreted as normal ranged from 39% to 74%, and the proportion assessed as abnormal ranged from 3% to 43%. All five observers agreed on the interpretation of 29% of the tracings. Inter-reviewer reproducibility scarcely changed when clinical data were provided. These findings emphasize the need to evaluate all methods of fetal monitoring before they become widespread.