Precision biased by region of interest and rotation of the hip
Background and purpose: Resurfacing Total Hip Arthroplasty (RTHA) may preserve the femoral neck bone-stock post-operatively. Bone Mineral Density (BMD), could theoretically be affected by the hip-position, and bias longitudinal studies. We aimed to investigate BMD precision dependency on type of ROI and position of hip. Method: We DXA scanned the femoral neck of 15 resurfacing patients twice with the hip in 3 different rotations; 15° internal, neutral, and 15° external. For each position BMD was analyzed with 3 different surface area models. One model measured BMD in the total femoral neck, the second model divided the neck in two and the third model had 6 divisions. Results: When all hip positions were pooled a mean Coefficient of variation (CV) of 3.1%, 3.6% and 4.6% was found in the 1, 2 and 6-region models, respectively, The external rotated hip position was less reproducible. When the hip was rotated in increments of 15° and 30°, the mean CVs rose to 7.2%, 7.3% and 11.8%. Rotation affected the precision most in the model that divided the neck in 6 sub regions, predominantly in the lateral and distal regions. For larger-region models, some rotation could be allowed without compromising the precision. Interpretation: If hip rotation is strictly controlled, DXA can reliably provide detailed topographical information about the BMD changes around a RTHA. As rotation strongly affects the precision of the BMD measurements in small regions, we suggest applying a less detailed analyzing model in studies where the leg position has not been firmly controlled.
Acta Orthopaedica Scandinavica, 2010, Vol 81, Issue 3, p. 318-323