• BACKGROUND
    • Many series have found that certain associated injuries occur with greater frequency in patients with scapula fractures than in patients without scapula fractures. However, several of the published series were limited by lack of a control group, inclusion of a patient population limited to the catchment area of one hospital, or inadequate control for injury severity. The goal of this study was to determine whether there was a relationship between scapula fractures and concomitant injury and which injuries related simply to the increased injury severity observed in this patient population.
  • METHODS
    • This series was a retrospective case control database analysis. Patients were identified through the National Trauma Database from 1994 to 2002 from trauma centers across the United States. Diagnosis code (ICD-9) 811.0 was used to identify 9,453 scapular fractures, whereas the 2,728 patients in the control group were selected by random number generation. After data extraction to a database, each patient was examined for concomitant diagnoses. The binomial distribution was used to compare cases and controls, as well as different diagnostic groups before adjusting for injury severity. The Bonferroni correction was applied to correct for the multiple null hypotheses. After univariate analysis, the data were analyzed with logistic regression using injury severity score as a covariate.
  • RESULTS
    • After statistical adjustment for multiple tests, there was not a statistically significant difference in injury rates for patients with different types of scapula fractures. However, there were many injuries that showed increased frequency in patients with scapula fractures compared with patients without this injury. Interestingly, when injury severity was statistically adjusted for, many of these differences disappeared.
  • CONCLUSIONS
    • After adjustment for injury severity, upper extremity, thoracic, and pelvic ring injuries were associated with greater frequency in patients with scapular fracture. The majority of other injuries found to occur frequently in the unadjusted patient population were likely because the injury severity is higher in patients with scapula fractures.
  • LEVEL OF EVIDENCE
    • III.