The treatment of acromioclavicular (AC) joint separations is controversial, particularly for Rockwood type III injuries. Rockwood type IV injuries, which correspond to horizontal instability, are very likely under-diagnosed. The objective of this study was to evaluate the inter- and intra-observer reproducibility of the Rockwood classification through an evaluation of standard radiographs, as described in the original article.

This was a prospective radiographic study using protocol-based data from the 2014 symposium of the French Society of Arthroscopy (SFA). Fifteen anonymized radiological records were analysed by six independent examiners on two occasions, 1 week apart. The records consisted of a comparative A/P view of the two acromioclavicular joints (Zanca view), an axillary lateral view and dynamic lateral views (Tauber protocol) to uncover dynamic horizontal instability. A detailed analysis protocol was implemented that included absolute and relative measurements on each view; the relative measurements were used to account for radiographic magnification.

The inter- and intra-observer reproducibility on the A/P radiographs was good to excellent. The reproducibility was fair to good on the lateral views, but the measurements varied greatly from one subject to another, and significant errors were found with certain records. The reproducibility of the dynamic views proposed by Tauber was poor to fair.

Radiographic analysis of AC joint separations is reproducible in the vertical plane, which makes it possible to diagnose Rockwood type II, III and V injuries. On the other hand, static and dynamic analyses in the horizontal plane do not have good reproducibility and do not contribute to make an accurate diagnosis of Rockwood type IV injuries.

Level I, Diagnostic study.