Table 2

Interobserver agreement for CT scores allocated to infants with NBS-diagnosed CF at 1 year of age during initial scoring of LCFC scans (n=65)

(a) Bronchial dilatation (maximum possible score=72)
κ=0.21 (0.05; 0.37)Scorer A
012345
Scorer B
 048
 16
 2431
 31
 41
 51
(b) Air trapping (maximum possible score=27)
κ=0.66 (0.49; 0.83)Scorer A
012345781516
Scorer B
0371
163
2311
31112
42
52
71
811
151
16
(c) Total CT score (maximum possible score=243)
κ=0.34 (0.20 to 0.49)Scorer A
01234567891012131417192530
Scorer B
07
155
271
3521
43121
54111
6111
71111
811
91
101
121
132
141
171
191
25
301
  • Shaded cells across the diagonals within each table represent identical results by scorers A and B using the Brody-II scoring system. For air trapping scores >5 and total CT scores >12, only those for which any values were obtained are shown. Although scorer B identified more abnormalities on scans than scorer A as indicated by values generally falling below the shaded diagonal cells (17 (26%) vs 5 (7%) for bronchial dilatation; 27 (42%) vs 17 (26%) for air trapping, the severity of changes were generally very minor, with only seven (11%) and two (3%) of patients having a total CT score ≥12 or 5% of the total possible score). κ=κ coefficient (95% CI) as a measure of agreement of CT subscores and total scores allocated by scorer A and B. It can be seen that the majority of discrepancies for bronchodilatation occurred when changes were deemed to be very minor(1–3) by one scorer and absent [0] by the other.

  • LCFC, London Cystic Fibrosis Collaboration; NBS, newborn screened.