Initial horizontal AP imaging showed completely displaced midshaft clavicle fractures in 9 of the 10 patients, and 15° simulated radiographs showed completely displaced fractures in all 10 patients (P = .50).
Discussion
Our study results demonstrated that an upright 15° radiographic tilt (AP cephalad or PA caudal) identified the most fracture displacement in the most patients with acute midshaft clavicle fractures. To our knowledge, this is the first study to identify the radiographic angulation that best shows the most clavicle fracture fragment displacement.
Other investigators have studied the accuracy of different radiographic views in the assessment of midshaft clavicle fractures, but they concentrated on fracture shortening. Smekal and colleagues9 used computed tomography (CT) and 3 different radiographic views to evaluate malunited midshaft clavicle fractures. Comparing the horizontal clavicular length measurements obtained with radiographs and CT scans, they determined that PA thoracic radiographs were in highest agreement with the CT scans. The results, however, were not statistically significant. In their study, supine CT was successful because the fractures were healed, and the displacement and shortening amounts were not affected by patient position. Sharr and Mohammed10 studied the accuracy of different views in the assessment of clavicle length in an articulated cadaver specimen. They obtained multiple AP and PA radiographs of different horizontal (medial, lateral) and vertical (cephalad, caudal) angulations. Actual clavicle length was then directly measured and compared with the length measured on the different views. The authors concluded that a PA 15° caudal radiograph was most accurate in assessing clavicular length. Both Smekal and colleagues9 and Sharr and Mohammed10 recommended the PA radiograph because it decreases the degree of magnification on AP radiographs by minimizing the film-to-object distance.
Our findings are important because more accurate determination of fracture displacement in patients with midshaft clavicle fractures may change clinical management. Nowak and colleagues11 investigated various patient and clavicle fracture characteristics that were predictive of a higher rate of long-term sequelae. They found that complete fracture displacement was the strongest radiographic predictor of patients’ beliefs that they were fully recovered from injury at final follow-up. The authors concluded that fractures with no bony contact should receive more “active” management. Robinson and colleagues12 studied a cohort of patients with nonoperatively managed midshaft clavicle fractures and concluded that complete fracture displacement significantly increased risk for nonunion (this risk was 2.3 times higher in patients with displaced fractures than in patients with nondisplaced fractures). Last, McKee and colleagues13 found that shoulder strength and endurance were significantly decreased in nonoperatively treated displaced midshaft clavicle fractures than in the same patients’ uninjured shoulders.
Extending the results of these studies, recent prospective randomized control trials and a meta-analysis have compared the clinical outcomes of nonoperatively and operatively managed displaced midshaft clavicle fractures.14-18 With few exceptions, these studies found improved clinical results with operative fixation. In one such study, the Canadian Orthopaedic Trauma Society14 randomized patients with displaced midshaft clavicle fractures to either operative plate fixation or sling immobilization. The operative group was found to have improved Disability of the Arm, Shoulder, and Hand scores, improved Constant shoulder scores, increased patient satisfaction, faster mean time to bony fracture union, higher satisfaction with shoulder appearance, and lower rates of nonunion and malunion. Given the results of these studies, accurate identification of a displaced midshaft clavicle fracture with no cortical contact is fundamental in deciding whether to recommend operative fixation.
Retrospective review of our cohort’s initial radiographs revealed 1 case in which the patient’s completely displaced midshaft clavicle fracture would not have been diagnosed solely with an AP horizontal image. Cortical contact was seen on a standard AP clavicle radiograph (Figures 2A, 2B), and a 15° tilt radiograph created from 3-D fluoroscopy scan showed complete fracture fragment displacement (Figure 3). A change in fracture classification from partially displaced to fully displaced could alter the type of management used by a treating surgeon.
There were obvious weaknesses to this study. First, its sample size was small (10 patients). Nevertheless, we had sufficient numbers to find a statistically significant angulation. Second, a wider range of radiographic angles could have been studied. Our intent, however, was to investigate the accuracy of the 2 most common supplementary clavicle views (20° and 45° cephalic tilt). Therefore, we selected a range of simulated radiographs that began 5° outside these angulations. Third, we measured only the degree of fracture displacement; we were unable to accurately access fracture shortening, as the 3-D fluoroscopic images were limited to the injured clavicles. A potential solution to this problem is to widen the exposure field in order to include the entire chest and allow clavicular length comparison against the uninjured side. Doing this would have been possible, but at the expense of increasing the patient’s radiation exposure.