Classification of postoperative acromial fractures following reverse shoulder arthroplasty.

BACKGROUND: Acromial fractures following reverse shoulder arthroplasty are difficult to identify and have variable effects on outcomes. A reproducible classification system is needed to categorize these fractures before it is possible to evaluate optimal treatment. The purposes of this study were to investigate the ability of radiographs to detect acromial fractures and establish fracture union, and to assess the reliability of a new classification system of postoperative acromial fractures based on the involvement of the deltoid origin.
METHODS: Eighteen patients presenting with pain along the acromion or scapular spine following reverse shoulder arthroplasty were evaluated for acromial fractures over a period of fifty-six months. If radiographic findings were negative, computed tomography (CT) scans were performed. Sixteen acromial fractures were identified and classified into subtypes on the basis of the deltoid muscle origin: type I indicated involvement of a portion of the anterior and middle deltoid origin; type II, at least the entire middle deltoid origin with a portion but not all of the posterior deltoid origin; and type III, the entire middle and posterior deltoid origin. The average age of the patients at the time of surgery was seventy-seven years (range, fifty-seven to eighty-seven years); the average duration of follow-up after surgery was twenty-five months (range, seven to forty-eight months). All fractures were treated nonoperatively. Blinded, fellowship-trained shoulder surgeons reviewed images that best illustrated the fracture, and the interobserver reliability of the classification system was evaluated. The reliability of radiographs to detect the fracture at the time of the initial presentation of pain and to observe ultimate fracture union was evaluated.
RESULTS: Interobserver reliability of the classification system was excellent (interclass correlation coefficient of 0.96). Radiographs were unreliable at detecting acromial fractures at the time of the initial presentation of pain (κ = -0.5) and at determining fracture union (κ = 0.05).
CONCLUSIONS: Radiographs may be unreliable for detecting acromial fractures after reverse total shoulder arthroplasty, and CT scans are often needed to identify the fracture. Once identified, postoperative acromial fractures can be reliably classified into three subtypes. Nonoperative treatment of postoperative acromial fractures results in limitation of functional outcomes. LEVEL OF EVIDENCE: Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.