Scapular Fractures in the Pan-scan Era

Abstract

Background: Scapular fractures have been traditionally taught to be associated with significant injuries and major morbidity. As we demonstrated with sternal fracture, pulmonary contusion, and rib fracture, increased chest computed tomography (CT) utilization and head-to-pelvis CT (pan-scan) protocols in blunt trauma evaluation, however, may diagnose minor, clinically irrelevant scapular fractures, possibly rendering previous teachings obsolete.

Objectives: The objectives were to determine the 1) percentages of scapular fractures seen on chest CT only (SOCTO) versus seen on both chest x-ray (CXR) and CT and of isolated scapular fracture (scapular fracture without other thoracic injuries); 2) frequencies of associated thoracic injury with scapular fracture; and 3) proportion of patients admitted, mortality, hospital length of stay, and injury severity scores (ISS), comparing four patient groups: scapular fracture, nonscapular fracture, scapular fracture SOCTO, and isolated scapular fracture.

Methods: We conducted a preplanned analysis of patients prospectively enrolled in the NEXUS Chest CT study at nine Level I trauma centers with the following inclusion criteria: age > 14 years, blunt trauma within 6 hours of ED presentation, and receiving chest imaging during ED trauma evaluation.

Results: Of 11,477 subjects, 4,501 (39.2%) patients who had both CXR and chest CT and 2.7% of these had scapular fractures; 60.3% of these were SOCTO and 23 (19.0%) were isolated scapular fracture. The most commonly associated thoracic injuries were rib fracture, pulmonary contusion, pneumothorax, and thoracic spine fracture and all injuries were more common in scapular fracture patients than nonscapular fracture patients. Although scapular fracture patients had higher admission rates (86.8% vs. 47.4%; difference in proportions = 39.4% [95% confidence interval {CI} = 32.8% to 44.1%]), ISS (21 vs. 5), and length of stay (9.2 days vs. 5.6 days; mean difference = 3.4 days [95% CI = 2.1 to 4.7 days]) than patients without scapular fracture, their hospital mortality was not significantly different (5.6% vs. 3.0%; difference in proportions = 2.6% [95% CI = -8.2% to 0.3%]; unadjusted odds ratio = 1.9 [95% CI = 0.9 to 4.2]). Patients with scapular fracture SOCTO and isolated scapular fracture had higher admission rates and median ISS than nonscapular fracture patients, but their mortality was similar.

Conclusions: Under current blunt trauma imaging protocols that commonly include chest CT, most scapular fractures are SOCTO and most are associated with other thoracic injuries. Although patients with scapular fracture SOCTO and isolated scapular fracture have higher admission rates and ISS than nonscapular fracture patients, their hospital mortality is similar.