Table 3 Cross‐sectional (on baseline measurements) and longitudinal (on baseline–M12 changes measurements) interobserver and intraobserver reproducibilities of the four scoring methods, using intra‐class coefficient of correlation and the Bland–Altman approach.
| Scoring method | Inter‐reader | Reproducibility | ICC on M0–M12 change (95% CI) reading 1, (%) | Reader | Intra‐reader | Reproducibility | ICC on M0–M12 change (95% CI) reading 1, (%) | ||
|---|---|---|---|---|---|---|---|---|---|
| ICC on baseline scorings (95% CI), (%) | Smallest detectable difference | ICC on baseline scorings (95% CI), (%) | Smallest detectable difference | ||||||
| Global score | 0.859 (0.819 to 0.899) | 19.8 | 0.999 (0.998 to 1) | Obs 1 | 0.922 (0.899 to 0.945) | 20.6 | 0.939 (0.921 to 0.957) | ||
| Obs 2 | 0.961 (0.949 to 0.973) | 13.6 | 0.956 (0.943 to 0.969) | ||||||
| Kellgren–Lawrence | 0.951 (0.936 to 0.966) | 14.2 | 0.998 (0.997 to 0.999) | Obs 1 | 0.988 (0.984 to 0.992) | 10.2 | 0.990 (0.980 to 1.000) | ||
| Obs 2 | 0.991 (0.988 to 0.994) | 6.6 | 0.998 (0.997 to 0.999) | ||||||
| Kallman | 0.706 (0.631 to 0.781) | 8.8 | 0.995 (0.993 to 0.997) | Obs 1 | 0.962 (0.950 to 0.974) | 7.8 | 0.986 (0.982 to 0.990) | ||
| Obs 2 | 0.999 (0.998 to 1.000) | 6.8 | 0.959 (0.947 to 0.971) | ||||||
| Verbruggen | 0.996 (0.994 to 0.998) | 13.4 | 0.998 (0.997 to 0.999) | Obs 1 | 0.999 (0.998 to 1.000) | 4 | 0.941 (0.922 to 0.958) | ||
| Obs 2 | 0.999 (0.998 to 1.000) | 5.2 | 0.988 (0.984 to 0.992) | ||||||
ICC, intra‐class coefficient of correlation; M, month; SDD, smallest detectable difference.
SDD was obtained from twice the SD of the mean of the differences.