INTRODUCTION: The present study was designed to examine whether oblique radiographs (Judet views) in addition to 2D and 3D CT scans improved the intra- and interobserver reliability when assessing acetabular fractures. MATERIALS AND METHODS: Four international orthopedic pelvic trauma centers reviewed the radiological images for 20 acetabular fracture patients. Three different image sets were made; one set containing plain radiographs including oblique (Judet) views and 2D axial CT scans. The second set contained an AP radiograph of the pelvis, without oblique views, 2D and 3D CT scans. The third set contained all the images. The image sets were evaluated in three separate sessions, for each session the raters were asked to classify the fracture according to Letournel, as well as record a number of other important radiological features concerning the fracture. RESULTS: The interobserver agreement for the Letournel classification was found to be moderate for all image sets. The image set without oblique views showed the best agreement with a kappa value of 0.60. The intra- and interobserver agreement for important modifiers were found to be substantial. The addition of oblique radiographs did not seem to increase the intra- or interobserver agreement for any of the factors evaluated except for the roof arc score. CONCLUSION: The moderate agreement found for the Letournel classification is to be expected given the complexity of the classification. The addition of oblique radiographs to the image sets does not seem to improve the reliability and thus its routine use for classification and decision making may be debated.
INTRODUCTION: The present study was designed to examine whether oblique radiographs (Judet views) in addition to 2D and 3D CT scans improved the intra- and interobserver reliability when assessing acetabular fractures. MATERIALS AND METHODS: Four international orthopedic pelvic trauma centers reviewed the radiological images for 20 acetabular fracturepatients. Three different image sets were made; one set containing plain radiographs including oblique (Judet) views and 2D axial CT scans. The second set contained an AP radiograph of the pelvis, without oblique views, 2D and 3D CT scans. The third set contained all the images. The image sets were evaluated in three separate sessions, for each session the raters were asked to classify the fracture according to Letournel, as well as record a number of other important radiological features concerning the fracture. RESULTS: The interobserver agreement for the Letournel classification was found to be moderate for all image sets. The image set without oblique views showed the best agreement with a kappa value of 0.60. The intra- and interobserver agreement for important modifiers were found to be substantial. The addition of oblique radiographs did not seem to increase the intra- or interobserver agreement for any of the factors evaluated except for the roof arc score. CONCLUSION: The moderate agreement found for the Letournel classification is to be expected given the complexity of the classification. The addition of oblique radiographs to the image sets does not seem to improve the reliability and thus its routine use for classification and decision making may be debated.