Can the crossover sign be a reliable marker of global retroversion of the acetabulum?

Diagnosis of acetabular retroversion based on crossover sign in the anteroposterior radiograph of the hip joint is well described. Accuracy of the crossover sign to identify global retroversion of the acetabulum in comparison to version of the acetabulum in reconstructed three-dimensional computed tomography (3D CT) scan of the hip was the aim of this study. X-rays of 500 hips were assessed regarding presence of crossover sign and its location in the upper, middle, or lower third of the acetabulum. Mean of anteversion and true retroversion (defined as less than one standard deviation below the mean of acetabular anteversion) of the acetabulum using reconstructed 3D CT scan by mathematical software was determined among 500 hips. The positive and negative crossover signs were compared to the retroversion obtained by CT scan. The average of acetabular anteversion was 12.5 ± 4.2 degrees. True global retroversion in 3D CT scans was defined as a version below 8.3 degrees. Although positive crossover sign was seen in 193 out of 500 (38 %), only 69 out of 500 (13.8 %) of hips had version below 8.3 (true retroversion) and 124 subjects had an acetabular version above 8.3. The sensitivity and specificity of crossover signs were about 82 and 70 %, respectively. The crossover sign could pick up hips with less than normal anteversion with acceptable sensitivity but it has no enough specificity for being used as the sole indication for treatment.