relative to the scapula on standardized plain radiographs

Deana Mercer, MDad, Matthew D. Saltzman, MDb, Moni Blazej Neradilek, MSc, Brian B. Gilmer, MDa, Winston J. Warme, MDa, Frederick A. Matsen III, MDa

^{a Department of Orthopedics and Sports Medicine, University of Washington, Seattle, WA, USA
b Department of Orthopaedic Surgery, Northwestern University, Chicago, IL, USA
c The Mountain-Whisper-Light Statistics, Seattle, WA, USA
d Department of Orthopaedics and Rehabilitation, University of New Mexico, Albuquerque, NM, USA}

Journal of Shoulder and Elbow Surgery, volume 20, issue 3, Pages 363-371 (March 2011)

Background Recent articles in this journal showed the clinical importance of the position of the humeral head center in relation to the glenoid. However, the precision, reproducibility, and sensitivity of this and other methods of documenting the head center position have not been evaluated in detail.

Materials and methods We used templates to fit a coordinate system to the scapular anatomy visible on standardized radiographs. Two observers then used these templates to measure the position of the head center relative to this coordinate system on 25 normal shoulder radiographs and on 25 radiographs of shoulders with cuff tear arthropathy (CTA).

Results Head center measurements had excellent precision. Normal shoulder radiographs showed a consistent head center position (0.7 ± 1.7 mm medial and 0.6 ± 1.3 mm inferior to the coordinate origin on the anteroposterior view and 0.1 ± 1.3 mm medial and 0.0 ± 1.3 mm anterior to the coordinate origin on the axillary view). The head center of CTA shoulder radiographs was 10.18 ± 5.16 mm above the coordinate origin on the anteroposterior view, significantly different from that for the normal shoulder radiographs (P < .001).

Discussion The relative position of the humeral head center to the scapula determines the resting length and the moment arms of the scapulohumeral muscles. Correlation of shoulder function with the head center position may provide insights into both shoulder pathomechanics and the optimization of shoulder arthroplasty.

Conclusion This practical technique showed a high degree of precision and reproducibility for normal and CTA shoulder radiographs as well as a high level of discrimination between these two groups.