Radiologic assessment of humeral head retroversion. Description of a new method

A radiologic method for assessment of the humeral head retroversion angle has been developed using one radiograph obtained in the semi-axial view. Validity and reliability of the method has been tested. In five healthy volunteers both shoulders were examined both with CT and with the new radiographic method. The average difference in angle determinations between the methods was 1.5 degrees and the maximum difference was 2 degrees. Angle determination on radiographs from 22 healthy shoulders was performed by two independent radiologists. The coefficient of variation for intraobserver measurements was 2.8 per cent and for interobserver measurements it was 4.6 per cent. Three isolated humerus bones were examined in multiple semi-axial projections and the humeral head retroversion was measured. The effect of humeral position (flexion, extension, abduction) on angle assessments was analyzed. A method error exceeding 2 degrees was only seen when the specimens were in an extended or extremely abducted position. It is concluded that with the arm in the correct position measurements of humeral head retroversion can be performed with this method with high accuracy.