Accuracy of in situ neck-shaft angle and shortening measurements of the anatomically reduced, varus malreduced and shortened proximal femur: can we believe what we see on the postoperative films?

Abstract

Measuring the neck-shaft angle (NSA) and amount of shortening of the femoral neck on the anterior to posterior (AP) X-ray is important when treating proximal femur fractures. To compensate for proximal femoral external rotation, the X-rays need to be taken with the leg internally rotated, an act that cannot always be performed or verified. This study aims to define the utility of in situ AP X-ray in NSA and shortening measurements. Computed tomography (CT) scans of 50 patients undergoing abdominal CT scans were assessed for the in situ rotation of the femoral neck relative to the AP beam. Three proximal femur fracture Sawbones models were made and AP X-rays of the models were taken with changing proximal femur rotation. NSA and shortening were measured on all X-rays. In situ femoral neck rotation averaged 25.4±10.6° of external rotation (range, 0.9-51.8°, 80% of measurements less than 35°). NSA measurements varied less than 5° with less than 35° of rotation in all models, and were always greater than the true value. Femoral neck vertical length (VL) measurement was independent of proximal femur rotation whereas the horizontal length component was found to be highly dependent on the same. NSA measured on AP X-ray will be accurate to within 5° in 80% of patients with the hip left in situ and in 100% of the patients if the hip is internally rotated 15°. Measurement of significant varus or loss of VL of the femoral neck can be considered accurate regardless of leg rotation at the time of X-rays being taken.