κ values with 95% confidence intervals for the three pairs of observers for the final diagnosis in the whole study population (n=108)
| κ values | |||
|---|---|---|---|
| A v B | A v C | B v C | |
| Final diagnosis (including the category “diagnosis not possible because of poor quality”) | 0.26 (0.16 to 0.36) n=99 | 0.41 (0.29 to 0.54) n=103 | 0.26 (0.16 to 0.35) n=98 |
| Final diagnosis possible/not possible because of poor quality | 0.11 (−0.01 to 0.22) n=98 | 0.39 (−0.16 to 0.94) n=102 | 0.07 (−0.02 to 0.17) n=96 |
| Final diagnosis (excluding cases without a definite diagnosis because of poor quality) | 0.39 (0.24 to 0.54) n=63 | 0.43 (0.30 to 0.56) n=98 | 0.40 (0.26 to 0.54) n=62 |
Results are given for all children seen by both observers. After exclusion of cases in which at least one of the pair felt the quality too poor to make a definite diagnosis, the difference between the three pairs vanished.