Validity, reliability and reproducibility of plain radiographic measurements after total hip arthroplasty

In total hip arthroplasty (THA), radiographic preoperative planning and postoperative evaluation of acetabular component, femoral offset (FO) and leg length discrepancy (LLD) require good validity, interobserver reliability and intraobserver reproducibility. In this study, we evaluated the validity of the Sundsvall method of FO measurement and the interobserver reliability and intraobserver reproducibility of measurement of FO, LLD, acetabular cup inclination and anteversion.

Ninety patients with primary unilateral osteoarthritis (OA) were included in this prospective study. On postoperative radiographs FO by the Sundsvall method (femoral axis-pelvic midline), FO by a standard method (femoral axis-hip rotational centre-teardrop point), LLD (inter-teardrop line-lesser trochanter), acetabular cup inclination (on AP view the angle between the cup rim and transischial line) and anteversion (on lateral view the angle between the face of acetabulum and a line perpendicular to the horizontal plane) were measured. The interobserver reliability and intraobserver reproducibility were calculated for three independent observers. The concurrent validity and degree of prediction of the Sundsvall method are measured by comparing its results with the standard method.

The interobserver reliability of all measurements was excellent (ICC >0.80), except for LLD, which was substantial (ICC = 0.79). The intraobserver reproducibility of all measurements was excellent (ICC > 0.80). The concurrent validity of the Sundsvall method compared to the standard method was good with a positive correlation.

The Sundsvall method is as reliable as the standard method. The evaluated radiographic measurement methods have the required validity and reliability to be used in clinical practice.