Abstract
The purpose of this study was to assess the Interobserver Reliability of the Sauvegrain Skeletal Age Assessment.
Elbow radiographs requested to exclude injury were anonymised. Sixteen examinations were assessed by ten independent orthopaedic specialist registrars or consultants. The Sauvegrain method as modified by Dimeglio was used to score the radiographs. The observations made were then assessed for interobserver reliability by means of a multiple observer Kappa score and the total scores by intra-class correlation coefficient.
Kappa scores for the components of the score were 0.403 for the lateral condyle; 0.492 for the trochlea; 0.354 for the proximal radius and 0.508 for the olecranon. Adding item scores to produce a modified Sauvegrain score had an intraclass reliability of 0.858 (95% CI 0.758 to 0.935).
Methods of identifying skeletal maturation and predicting future growth generally depend on the use of an atlas of hand radiographs. Difficulties with poor interobserver reliability associated with these methods has led to a move towards assessments that do not depend upon bone age estimations. Unfortunately plans based on ratios of growth or average patterns produce errors when unusual types of growth disturbance are present. We conclude that use of a scoring system for maturation assessed by elbow radiographs offers a significant advantage when substituted into the straight line method of growth prediction. The Sauvegrain method as modified by Dimeglio1 has demonstrated an excellent level of inter observer reliability. We have used Sauvegrain scores to improve the accuracy of timing when using the Mosely straight line method.
Correspondence should be addressed to Editorial Secretary Mr ML Costa or Assistant Editorial Secretary Mr B.J. Ollivere at BOA, 35–43 Lincoln’s Inn Fields, London WC2A 3PE, England; Email: mattcosta@hotmail.com or ben@ollivere.co.uk
