AP PELVIC RADIOGRAPHS–IS WHAT YOU SEE WHAT YOU GET?

Abstract

Introduction: Recently, the correct interpretation of anteroposterior (AP) pelvic radiographs has regained increased attention, particularly in the field of joint preserving hip surgery. The diagnosis of acetabular retroversion associated with femoroacetabular impingement or hip dysplasia is made regardless the individual pelvic orientation due to the lack of a method of correction. Furthermore, it is known that a substantial number of the most common radiographical hip parameters can vary with the individual pelvic orientation. The goal of the study was to evaluate which parameter can be measured accurately on an AP radiograph.

Methods: Digital AP pelvic radiographs of 100 consecutive hips were used for evaluation. The blinded and randomized x-rays were examined by two independent observers with special software that has been validated previously. The software is able to correct the projected acetabular rim and the associated parameters for pelvic malpositioning. The following parameters were investigated: femoral head coverage in craniocaudal and anteroposterior direction (in total and for each single quadrant of the femoral head), the lateral center edge angle, the acetabular index, the ACM-angle, the extrusion index, the cross-over sign, the retroversion index, and the posterior wall sign. All parameters were first measured regardless to the individual tilt and rotation. These non-standardized values were then compared to the standardized values for a neutral pelvic orientation. This was defined with a pelvic inclination of 60 degrees which was detected with one single strong lateral pelvic radiograph.

Results: There were no differences in evaluation of the radiographs between the two observers concerning the significance of standardized and non-standardized values for the measured features. All but three parameters were significantly different when measured to the anatomically reference neutral orientation. The only parameters that did not change after standardization were the total femoral coverage, the acetabular index and the ACM.

Discussion: Except from the ACM and the acetabular index, basically all parameters change when standardized to a neutral orientation. Although from a statistical point of view, the total craniocaudal femoral coverage did not change, it is likely that this is due to an inverse effect of the anterior and posterior part of the acetabulum. We conclude that the most common hip parameters can not be reliably measured without standardization. It remains to be proven that the standardization of the parameters correlates with the clinical symptoms.