Daniela del Río-González, Servicio de Traumatología y Ortopedia, Centro Médico ABC, Ciudad de México, México
Eira Santiago-Balmaseda, Servicio de Traumatología y Ortopedia, Centro Médico ABC, Ciudad de México, México
Luis Ochoa-Olvera, Servicio de Traumatología y Ortopedia, Centro Médico ABC, Ciudad de México, México
Carlos F. Gutiérrez-Toledano, Servicio de Traumatología y Ortopedia, Centro Médico ABC, Ciudad de México, México
Jorge A. Ascencio-Aragón, Servicio de Traumatología y Ortopedia, Centro Médico ABC, Ciudad de México, México
Alejandro G. Gallegos-Tejeda, Servicio de Traumatología y Ortopedia, Centro Médico ABC, Ciudad de México, México
Background: Currently, the reproducibility and reliability of radiological classifications for proximal humerus fractures have not been demonstrated. Despite the widespread use of the Neer classification, it exhibits high interobserver variability. Objective: To assess interobserver variability of three radiographic classifications for proximal humerus fractures (Neer, Hertel, and Mayo-FJD) among orthopedic residents at Centro Médico ABC. Method: Four observers evaluated 100 anteroposterior radiographs of proximal humerus fractures, assigning a grade based on Neer, Hertel, and Mayo-FJD classifications. Interobserver agreement was measured using the kappa correlation coefficient. Results: Of the total radiographs, 70% were from female patients and 30% from male patients, with an average age of 67 years. The Neer classification showed the highest interobserver agreement (0.350). Hertel classification had an agreement of 0.265, and Mayo-FJD classification showed an agreement of 0.345, with the latter being more consistent among higher-ranking residents. Conclusions: The evaluated classifications lack reliability in describing proximal humerus fractures in anteroposterior shoulder radiographs. It is recommended to assess the reliability of these classifications for computed tomography studies.
Keywords: Classification. Proximal humerus. Fracture.