Related Experiment Video
Updated: Jul 16, 2026
Reverse Total Shoulder Arthroplasty
Published on: July 5, 2011
Diagnostic and treatment differences among experienced shoulder surgeons for instability conditions of the shoulder
Jas Chahal1, Kamrouz Kassiri, Anna Dion
1University of Manitoba, Faculty of Medicine, St. Boniface Hospital, Winnipeg, Manitoba, Canada.
Objective:
To quantify interobserver agreement in the diagnosis and treatment of shoulder instabilities among expert North American shoulder surgeons. We hypothesized that interobserver consistency among this group will be significantly low in both diagnosis and treatment.
Design:
Survey/Descriptive Epidemiology Study.
Setting:
Self-administered survey via e-mail.
Participants:
Active members of American Shoulder and Elbow Surgeons (ASES) and JOINTS Canada (Joined Orthopaedic Initiatives for National Trials of the Shoulder), whose practices consisted primarily of shoulder surgery.
Interventions:
Participants were sent a self-administered survey via e-mail and polled as to their choice of diagnosis and treatment in 5 different shoulder conditions.
Main Outcome Measurements:
A Kappa coefficient of agreement, Ksc, was used to measure relative interobserver reliability.
Results:
Overall response rate was 62.7% (42/67 surveys). The level of interobserver reliability was fair (Ksc 0.38, P < 0.0001) to almost perfect (Ksc 0.97, P < 0.0001) in diagnosing shoulder instability and slight (Ksc 0.23, P < 0.0001) to substantial (Ksc 0.69, P < 0.0001) for therapeutic approach. The greatest diagnostic differences were noted for a painful shoulder in a throwing athlete with subtle anterior instability (Ksc 0.43, P < 0.0001) and for a patient with voluntary posterior instability with an asymptomatic sulcus sign (Ksc 0.38, P < 0.0001). The greatest differences in treatment choice were for the throwing athlete with subtle anterior instability (Ksc 0.38, P < 0.0001), a patient with voluntary posterior instability (Ksc 0.34, P < 0.0001), and a patient with bidirectional instability (Ksc 0.23, P < 0.0001).
Conclusions:
These inconsistencies highlight the need for greater awareness and standardization of diagnostic criteria. This work may serve as the foundation for more universal treatment plans and subsequently more meaningful clinical outcomes.