Femoroacetabular impingement: can the alpha angle be estimated?

Femoroacetabular impingement is an important entity with well-described radiographic findings. One of the criteria of the cam type of femoroacetabular impingement is femoral head-neck dysplasia, denoted mathematically as the "alpha angle." Several observers have reported that direct measurement of the angle may not be necessary because subjective appraisal may yield similar results. We sought to scientifically determine the accuracy of a subjective assessment, using the calculated angle as the gold standard.

At 1.5 T, 50 consecutive patients' hips were evaluated on sets of oblique axial images. Two musculoskeletal radiologists recorded their subjective opinion as to the alpha angle using a confidence scale of 1-5. Direct mathematic measurement of the alpha angle was done by a third independent observer and correlated with the subjective results. Correlations between the subjective and measured angles and interobserver variation were calculated.

Statistically, significant variability was seen in the subjective assessment of the alpha angle. When the alpha angle was > 55 degrees, the area under the receiver operating characteristic curve (AUC) was 0.606, indicating that visual assessment is a poor predictor of a wide alpha angle. Even in patients with a measured normal alpha angle (< 55 degrees), slightly fewer than half were subjectively thought to possibly, likely, or definitely have abnormal angles. Similarly, more than half of the abnormal cases (alpha angles > 55 degrees) were subjectively thought to possibly or probably be normal.

Subjective assessment of alpha angles is suboptimal unless one is quite confident of a bone abnormality.