|| Checking for direct PDF access through Ovid
To determine and test mathematical models of the relationship between hip flexion–extension, femoral rotation, and NSA. We hypothesized that hip flexion–extension and femoral rotation would result in NSA measurement error.Two mathematical models were developed to predict NSA in varying degrees of hip flexion–extension and femoral rotation. The predictions of the equations were tested in vitro using a model that varied hip flexion–extension while keeping rotation constant, and vice versa. The NSA was measured from an anterior–posterior radiograph obtained with a C-arm. Attributable measurement error based on hip positioning was calculated from the models.The predictions of the model correlated well with the experimental data (correlation coefficient = 0.82–0.90). A wide range of patient positioning was found to result in <5–10 degrees error in the measurement of NSA. Hip flexion–extension and femoral rotation had a synergistic effect in measurement error of the NSA. Measurement error was minimized when hip flexion–extension was within 10 degrees of neutral.This study demonstrates that hip flexion–extension and femoral rotation significantly affect the measurement of the NSA. To avoid inadvertently fixing the proximal femur in varus or valgus, the hip should be positioned within 10 degrees of neutral flexion–extension with respect to the C-arm to minimize positional measurement error.