[Pelvic incidence reliability in spine sagittal balance]

The pelvic incidence (PI) is an anatomic parameter independent of the spatial orientation of the pelvis. However, measuring the PI remains a technically difficult task. PI is determined from radiographic construction which has several consequences. Imperfect radiographic incidence may compromise optimal conditions for measurement. Anatomic remodeling of the sacral plate can also have an impact on rigorous measurement of the PI. Several solutions have been proposed in the literature in order to obtain equivalent measurements. The proposed methods (Duval-Beaupere, Vialle) appear logical, but remain to be validated. The purpose of this study was to determine whether the angle measured on the plain x-ray is affected by the radiographic quality of the image and to determine whether the "variant pelvic incidence" (VPI) using a reference constituted by the posterior aspect of the sacrum, can be accepted as a valid equivalent to PI in the event of anatomic remodeling.

This was a prospective study conducted in 50 patients. The PI and the VPI were measured on 3D computed tomography reconstructions which allowed three series of measurements: in a strictly lateral view, in a view with the position of the femoral heads shifted in the horizontal plane, and a view with the position of the femoral heads shifted in the vertical plane. Statistical analysis was applied to compare the three series of measures in order to search for a significant difference for the two angles under consideration (PI and VPI) dependent on the position of the pelvis.

There was no significant difference in PI as a function of the position of the femoral heads: the coefficients of variation were all greater than 0.98 (p<0.01). For VPI, there was a significant difference between the theoretical ideal position and the horizontal shift of the femoral heads (p<0.05). Average measurements of PI and VPI were statistically different for each individual (p<0.05), with no identifiable relationship between the two values (PI and VPI).

Pelvic incidence is a reliable measure, even if the quality of the x-ray is not perfect. Conversely, the variant, using the posterior aspect of the sacrum, is highly dependent on pelvic orientation, with significant differences with position. We were unable to establish a relation between PI and VPI. In practice, it would be licit to consider that the straight line between the two femoral heads indicates the center of the femoral heads. On the contrary, it would not be wise to consider the posterior aspect of the sacrum as a reliable reference.