The axillary radiograph is part of the standard shoulder series after injury and is useful in assessing fracture anatomy and glenohumeral joint dislocations. The treatment of surgical neck fractures is based on angulation at the fracture site. Our hypothesis is that angulation seen on the initial axillary view may not correlate with angulation at union in the lateral plane as the radiograph is taken with the humerus in a position that is different than the treated position. The purpose of this study was to evaluate the correlation of the angulation on initial and healed radiographs of surgical neck fractures treated nonoperatively
Over a 5-year period, 36 patients with displaced isolated surgical neck fractures that were treated to union nonoperatively and had appropriate radiographs were identified and formed the basis of this report. Fracture angulation of the injury and healed AP, Y, and axillary views were measured by a single reviewer.
Using the Interclass Correlation Coefficient, there was moderate agreement between the injury and healed Y views (ICC 0.5, 95% CI: .16-.73), but only fair correlation between the injury and healed axillary views (ICC 0.35, 95% CI: -.09 to -.7). The poorest correlation was seen between the injury axillary and the healed Y view (ICC 0.08, 95% CI: -.1 to -.32).
The initial axillary view is thus confirmed to be a less predictive measure of the lateral angulation at union than the initial Y view. Axillary radiographs of surgical neck fractures taken at the time of injury may overestimate the angulation in the lateral plane that will be present if fractures are treated nonoperatively. The initial axillary radiograph should not be used to determine treatment of isolated displaced surgical neck fractures. The Y view is more predictive of angulation at union.