Table 3

 Reliability of acoustic analysis v stethoscope

WheezeRattlesCrackles
YesNoTotalYesNoTotalYesNoTotal
2×2 tables showing for how many infants there was agreement between the two measures; κ (95% confidence interval) given for each comparison. Rows and columns are defined as (A) and (B) observer 1 v observer 2; (C) and (D) occasion 1 v occasion 2.
A Between-observer agreement for stethoscope
Yes13142710414325
No2794182232831
Total15213614223663036
κ = 0.18 (−0.08 to 0.44)κ = 0.53 (0.21 to 0.86)κ = 0.46 (0.14 to 0.79)
B Between-observer agreement for acoustic analysis
Yes47117613448
No18925711112
Total515209112051520
κ = 0.24 (−0.12 to 0.60)κ = 0.22 (−0.18 to 0.63)κ = 0.44 (0.03 to 0.86)
C Within-observer agreement for acoustic analysis (instrument)
Yes299384444815722
No31720371043236
Total322658471158193958
κ = 0.57 (0.32 to 0.82)κ = 0.59 (0.34 to 0.85)κ = 0.56 (0.30 to 0.82)
D Within-observer agreement for acoustic analysis (repeat analysis)
Yes911016016617
No18912311112
Total109191721971219
κ = 0.79 (0.43 to 1.24)κ = 0.77 (0.33 to 1.20)κ = 0.77 (0.32 to 1.22)