Plain radiographs fail to reflect femoral offset in total hip arthroplasty.

Successful biomechanical reconstruction is a major goal in total hip arthroplasty (THA). We measured leg length (LL), global (GO) and femoral offset (FO) change on anteroposterior pelvis radiographs and on three-dimensional computed-tomography (3D-CT) with fiducial landmarks after cementless THA on 18 hips of cadaveric specimens. Measurements on radiographs were performed twice by four examiners and showed high interobserver (mean CCC ≥0.79) and intraobserver agreements (mean ICC ≥0.88). Mean differences between radiographic and 3D-CT measurements were 1.0 (SD 2.0) mm for LL, 0.6 (SD 3.6) mm for GO and 1.4 (SD 5.2) mm for FO. 1% of radiographic LL-, 15% of GO- and 35% of FO measurements were outside a tolerance limit of 5mm. Radiographs seem acceptable for measuring LL/GO change but fail to reflect FO change in THA.