Experienced epileptologists have very high confidence in their EEG interpretations and low to moderate I&IR, a common paradox in clinical medicine. Variance due to subjects (i.e., EEGs) was 65.3%, due to readers was 3.9%, and due to the interaction between readers and subjects was 30.8%.
![eeg interpretation eeg interpretation](https://images-na.ssl-images-amazon.com/images/I/71mB1Y3rLOL.jpg)
Cohen's kappa was not significantly different across rater pairs (chi-square = 17.3, df = 14, p = 0.24). Cohen's kappa ranged from 0.29 to 0.62 for the 15 reader pairs, with an aggregated Fleiss kappa of 0.44 for interrater agreement. Intrarater Cohen's kappa (κ c) ranged from 0.33 to 0.73 with an aggregated value of 0.59. Five of the six readers had a median confidence of ≥ 99%, and the upper quartile of confidence values was 100% for all six readers. A generalizability study assessed the contribution of subjects, readers, and the interaction between subjects and readers to interpretation variance. Each reader interpreted 150 unique studies, and 50 studies were re-interpreted to generate intrarater data. During two distinct time intervals, six board-certified clinical neurophysiologists classified 300 EEGs into one or more of seven diagnostic categories and assigned a subjective confidence to their interpretations. We measured both the intrarater reliability and the interrater reliability of EEG interpretation based on the interpretation of complete EEGs into standard diagnostic categories and rater confidence in their interpretations and investigated sources of variance in EEG interpretations.
![eeg interpretation eeg interpretation](https://image.slidesharecdn.com/eegncvemg-181104072043/85/eeg-interpretation-35-320.jpg)
The intrarater and interrater reliability (I&IR) of EEG interpretation has significant implications for the value of EEG as a diagnostic tool.