Computed tomography as a predictor of hip stability status in posterior wall fractures of the acetabulum.

OBJECTIVES: Studies using 2-dimensional computed tomography-derived criteria indicate that, in general, posterior wall fractures involving less than 20% of the posterior wall are stable and able to withstand physiologic loads, whereas those involving greater than 40%-50% are unstable, leaving a wide range of posterior wall fractures classified as indeterminate. The purpose of this study was to assess the ability of static measurement of posterior acetabular wall fragment size using computed tomography to predict hip stability status, as determined by dynamic stress examination under anesthesia.
DESIGN: Diagnostic level I. Retrospective analysis with testing of previously developed diagnostic criteria in a series of consecutive patients (with universally applied reference "gold" standard).
SETTING: Level I trauma center. PATIENTS: Thirty-three consecutive patients with isolated unilateral posterior wall (OTA 62-A1) acetabular fractures were evaluated by dynamic fluoroscopic stress testing under general anesthesia (examination under anesthesia) to determine hip stability status and subsequent clinical treatment. INTERVENTION: Three methods were used in a blinded fashion to calculate posterior wall fracture fragment size using 2-dimensional computed tomograms. These methods include those previously described by Calkins et al, which measures the smallest amount of intact acetabular arc, and Keith et al, which measures fragment size at the level of the fovea, and an alternative modification of that of Keith et al using the level of largest posterior wall deficit. Each method classifies hip instability into 3 groups: (1) stable, (2) indeterminate, and (3) unstable. The examination under anesthesia served as the gold standard. MAIN OUTCOME MEASUREMENT: Examination under anesthesia
RESULTS: Examination under anesthesia determined 15 hips to be unstable and 18 hips to be stable. The analyses showed that the methods of Calkins et al and Keith et al had a substantial percentage of incorrect predictions, especially in the critical group 1 patients (those predicted to be stable but were actually unstable). The percent incorrectly predicted for these group 1 patients was 33.3% (positive predictive value 66.7%) for the data derived form Calkins et al and 14.3% (positive predictive value 85.7%) for the data derived form Keith et al. In contradistinction, for the alternative method, specificity, sensitivity, and positive predictive value were all 100% with a 0% incorrectly predicted. However, with this alternative method, there was an increase in the number of group 2 fractures (23), as compared with the group 2 numbers for Calkins et al (n = 7) and Keith et al (n = 18). Reanalysis of the data for better potential cut points indicated that none of the methods could be improved in this way.
CONCLUSIONS: The alternative method is the only reliable technique that is predictive of hip stability for small fracture fragments while also being predictive of instability for large fracture fragments. However, these findings are based on small patient numbers, and there remain a substantial number of fractures involving 20% or more of the posterior wall that are both stable and unstable by examination under anesthesia. Therefore, given the low risk of the stress examination and the inherent problems making the computed tomography measurements, dynamic fluoroscopic stress testing under general anesthesia should be the preferred method for the determination of hip stability status after posterior wall fractures of the acetabulum.