Objective: A precise understanding of the radiological anatomy and biomechanics as well as reliable reference values of the hip are essential. The primary goal of this study was to provide reference values of the neck-shaft angle (NSA) for adult patients based on the analysis of rotation corrected computed tomography (CT) scans of 800 hips. The secondary aim was to compare these measurements with simulated anteroposterior roentgenograms of the pelvis.
Materials and methods: Pelvic CT scans of 400 patients (54.3 years, range 18-100 years; 200 female) were reconstructed in the derotated coronal plane of the proximal femur and as CT-based simulated anteroposterior roentgenograms of the pelvis in the anterior pelvic plane. Femora were categorized as coxa vara (<120°), physiologic (≥120° to <135°), and coxa valga (≥135°). Intra- and inter-rater reliability were analyzed.
Results: Primary research question: Mean NSA for male adults was 129.6° (range 113.2°-148.2°; SD 5.9°) and 131.9° (range 107.1°-151.9°; SD 6.8°) for females in derotated coronal reconstructions. Age (p < 0.001 in both views) and sex influenced the NSA significantly (p = 0.002 and p < 0.001); no significant differences were found between sides (p = 0.722 and p = 0.955). Overall, an excellent reliability of repeated measurements of one or two observers was found (ICC 0.891-0.995). Secondary research question: NSA values measured in the simulated anteroposterior roentgenogram and the rotation corrected coronal reconstruction differed significantly (p < 0.001).
Conclusions: While anteroposterior pelvis radiographs are susceptible to rotational errors, the coronal reconstruction of the proximal femur in the femoral neck plane allows the correct measurement of the NSA.
Keywords: Computed tomography; Hip biomechanics; Neck-shaft angle; Reference value; Total hip arthroplasty.