ABSTRACT
Objective:To investigate the reliability and repeatability of 3 classification systems for coronal fractures of distal humerus.Methods:Included were the imaging data (X-ray and CT scan) of the 64 patients with distal humeral coronal fracture who had been admitted to the outpatient and emergency departments of Beijing Jishuitan Hospital from January, 2015 to December, 2021. Four orthopedic surgeons (1 chief surgeon, 1 associate chief surgeon, 1 attending surgeon, and 1 resident surgeon) served as observers who classified the fractures by their imaging data respectively according to the Bryan-Morrey, Ring and Dubberley classifications. Four weeks later when the imaging data of 64 patients were renumbered, the 4 orthopedic surgeons performed the same classifications again. Inter-observer and intra-observer Kappa values were calculated for the 3 classifications. Landis and Koch Kappa's Benchmark Scale was used to interpret the results.Results:Interobserver confidence: The average agreement of the Bryan-Morrey classification was 75.0%, and its Kappa value was 0.546, indicating moderate agreement. The average consistency of Ring classification was 74.7%, and its Kappa value was 0.702, indicating high consistency. The average consistency of Dubberley classification was 84.1%, and its Kappa value was 0.785, indicating high consistency. Intra-observer repeatability: The average consistency of Bryan-Morrey classification was 89.9% and its Kappa value was 0.824 for the same observer. The average consistency of Ring classification was 91.0%, and its Kappa value was 0.884. The average consistency of Dubberley classification was 91.8%, and its Kappa value was 0.888.Conclusions:The 3 classification systems for distal humeral coronal fractures all demonstrate high reproducibility. As the Dubberley classification system is relatively more reliable, it has a certain guiding value for the treatment of coronal fractures of the distal humerus. However, in clinical practice, an optimal surgical strategy should take into account various factors that may affect fracture typing.