Back to AORI list of
AAOS presentations

Back to Fellowship News

Home | About AORI | Hip Replacements| Knee Arthroplasty AORI Physicians

The Accuracy & Reproducibility of Radiographically Assessing Stress Shielding: A Postmortem Analysis

C. Anderson Engh, Jr., MD, James P. McAuley, MD, Christi M. Sychterz, MS, Marie E. Sacco, BS, Charles A. Engh, MD

Based on this study, we question evaluating periprosthetic bone resorption by radiographic methods, since bone loss is not reproducibly recognized until 70% is lost.

This study assessed the ability of orthopaedic surgeons to reproducibly recognize changes in periprosthetic bone density from radiographs.

Materials & Methods: Twenty-nine unilateral hip replacements and the surrounding bone were retrieved at autopsy and radiographed; the contralateral normal femur was used as a control. Three surgeons independently examined specimen radiographs and classified bone loss in each of 16 femoral zones. Bone loss was considered present if the bone of the in vivo implanted femur showed evidence of cortical thinning, increased porosity, or decreased density when compared to the control femur. The kappa coefficient quantified inter- and intraobserver reproducibility. For 14 femoral pairs, we also quantified bone loss via dual energy x-ray absorptiometry (DEXA).

Results: We found that the surgeons agreed on the presence or absence of bone loss for 73% of the zones. The interobserver kappa value of .058 denoted only good reproducibility. The intraobserver reproducibility was better; one surgeon's initial evaluation of bone loss agreed with his second evaluation for 90% of the zones (kappa = 0.74).

DEXA analysis revealed that although some agreement between reviewers occurred when there was 20-60% bone loss, excellent agreement was not obtained until there was a mean 70% bone loss.

Conclusion: Based on these results, we suggest caution in interpreting plain radiographic analyses of femoral bone loss that do not provide interobserver reliability data. We question the utility of evaluating periprosthetic bone loss with radiographs, since it is not reproducibly recognized until 70% of the bone is gone.

Hip
Research