Figure 1.

Interobserver agreement between initial and rescoring London Cystic Fibrosis Collaboration (LCFC) rounds. CT scores were allocated by scorers A and B for the subset of 22 scans during initial and rescoring rounds. While all 22 pairs of results have been plotted, overlap of some data, particularly those with zero scores means that not all results can be identified individually. Bold circles represent data that overlaid each other, the number in brackets representing the number of infants with each combination of scores. During initial scoring of the subset, scores allocated by scorer B were generally higher than those by scorer A for bronchial dilatation and total scores (A and C). More consistent scores with good agreement were seen for air trapping (B). During rescoring of this subset, scores were more similar, although only fair agreement was again seen for bronchial dilatation (D), while good agreement was seen for air trapping and total scores (E and F). κ=κ coefficient (95% CI). AT, air trapping, BD, bronchial dilatation.