Agreement study of radiographic classification of rotator cuff tear arthropathy

This study evaluated the intra-rater and inter-rater correlation of 3 commonly used x-ray image classifications and defined the clinical factors most correlated with a surgical recommendation for a hemiarthroplasty or a reverse total shoulder arthroplasty (RSA) for treatment of rotator cuff tear arthropathy (CTA). We hypothesized that specific radiographic criteria and clinical criteria would be most important and consistently used among experienced shoulder surgeons when determining the best surgical option for a particular patient.

Four experienced orthopedic surgeons evaluated standard anteroposterior radiographs and the clinical examination of 37 shoulders with CTA. On each reading, they classified the grade of pathology using the Seebauer, Favard, and Hamada classifications. Using radiographic criteria alone, or with the clinical findings, each evaluator determined the recommended prosthetic treatment for each shoulder.

Intra-rater correlations for surgical recommendations using radiographic criteria ranged from 0.39 to 1.0 and improved in 3 of 4 evaluators when the clinical examination was included in the clinical decision. The inter-rater reliability using these same criteria were fair, at 0.32 for radiographic and .35 for radiographic and clinical data. The most significant radiographic factors associated with a surgical decision were the degree of humeral head superior migration and the escape of the humeral head from the coracoacromial arch. Clinical factors most associated with the decision for RSA were advanced age, loss of shoulder elevation, superior humeral head escape, and pseudoparalysis of the shoulder. Radiographic findings had a less significant effect on surgical recommendations when clinical factors were included.

Clinical and radiographic criteria are needed for a decision for hemiarthroplasty or RSA in the treatment of CTA. A treatment algorithm based upon radiographic and clinical criteria is proposed.