Accuracy of the axillary projection to determine fracture angulation of the proximal humerus

The accuracy of measuring angulation of stable proximal humerus fractures using the axillary lateral projection was investigated. A closing wedge osteotomy with apex anterior angulation was performed on two cadaveric humeri to simulate a stable surgical neck fracture. One specimen was fixed at 30 degrees angulation and the other at 55 degrees. Axillary radiographs were taken with each specimen articulating with the glenoid of a cadaveric scapula. The humerus was held in neutral rotation. Abduction was set at 30 degrees, 60 degrees, and 90 degrees. In each position of abduction, an axillary lateral radiograph was taken in 30 degrees forward flexion, neutral, and 30 degrees extension to simulate various arm positions. A total of nine radiographs were taken for each specimen. The axillary view is not accurate for measurement of proximal humerus angulation at the arm positions commonly encountered in the trauma setting.