Reproducibility of the Gartland classification for supracondylar humeral fractures in children.

PURPOSE To evaluate the intra- and inter-observer reproducibility of the Gartland radiographic classification for supracondylar humeral fractures in children. METHODS Anteroposterior and lateral radiographs of 50 supracondylar humeral fractures in children were graded on 2 separate occasions by 4 orthopaedic surgeons according to the Wilkins modification of the Gartland classification. Data were analysed by calculating the Kappa values for intra- and inter-observer agreement to indicate the reproducibility of the classification. RESULTS There was moderate inter-observer agreement, except for poor agreement over type I fractures. Type II fractures only showed fair to moderate agreement. Type III fractures and the flexion group showed good to very good agreement. Intra-observer agreement was good to very good. CONCLUSION Surgeons should treat paediatric supracondylar humeral fractures based on an assessment of the degree of displacement rather than by employing the Gartland classification.