Acute glenoid rim fractures: reliability of linear and surface area methods for CT-based bone loss assessment.

OBJECTIVE

We wanted to investigate and compare the reliability of two measuring methods for glenoid bone loss, one linear- and one surface area method, in acute glenoid fractures after dislocation. This was an experimental study.

METHOD: Thirty CT-shoulder scans in patients with an acute glenoid rim fracture were measured in a lateral"en face" view in 2D and 3D by two independent assessors. In 10 of 30 patients contralateral shoulder CT-scan was obtained for reference circle placement. Intra- and interrater reliability was analysed by Intra Class Correlation. Wilcoxon signed rank test was performed for comparison of bone loss estimates between methods.

RESULTS

The reliability ranged from 0.85 to 0.97 for the two methods in both 2D and 3D, which is considered good to excellent. The bone loss estimates were significantly larger with linear method than area method. Drawing of a best-fitting-circle had a good reliability between assessors.

CONCLUSIONS

Linear- and surface area methods demonstrated comparable reliability in assessing glenoid bone loss in acute fractures and we found no difference between 2D- and 3D-modalities. However, linear methods may overestimate defect in cases with irregular fracture lines, suggesting area method may be preferable in acute fractures.