Repeatability of sonographic measurements in clubfeet

To assess the intra- and interobserver agreement of ultrasound assessment of clubfoot patho-anatomy in early childhood.

Seventeen clubfeet in 12 children were sequentially scanned by 2 examiners and repeat assessments were carried out independently. Three well-defined imaging planes were chosen to evaluate navicular displacement: the medial malleolus-navicular (MM-N) distance, soft tissue thickness, talar deformity and the calcaneo-cuboid (C-C) distance. Intra- and interobserver agreement was analysed using Cohen's kappa for the discrete variables and by Bland-Altman's graphic technique for measurements.

Kappa values for intra-observer agreement were 0.82 for navicular displacement, 0.93 for "talar head pointing laterally", and 0.70 for medial deviation of the talar neck. The corresponding interobserver kappa values were 0.70, 0.68, and 0.36. The mean difference between the two observers for the MM-N distance was 0.42 +/- 3.0 mm and for the soft tissue thickness 0.35 +/- 2.0 mm; the C-C distance showed a mean interobserver distance of 0.0 +/- 2.8 mm.

The imaging planes used to study the talo-navicular and calcaneo-cuboid relationships are reproducible and relatively easy to learn. Intra- and interobserver assessments were acceptable for MM-N distance, soft tissue thickness, navicular displacement and "talar head pointing laterally", but questionable for the C-C distance and medial deviation of the talar neck.