Fluoroscopically assisted radiographs theoretically improve detection of total knee arthroplasty (TKA) implant loosening by providing a better evaluation of the true implant interface, but their utility has not been well studied. We sought to determine whether fluoroscopically guided radiographs improve the sensitivity, specificity, and interobserver reliability of determining TKA implant loosening compared to standard radiographs.
Standard anteroposterior and lateral and fluoroscopically assisted radiographs were retrospectively obtained from 60 patients within 6 months before revision TKA. Thirty knees were revised for aseptic loosening and 30 knees for other indications, most commonly instability. The radiographs were randomized. Four reviewers independently determined whether each tibial and femoral component was radiographically loose or stable. Intraoperative determination of implant stability was utilized as the gold standard.
Fluoroscopically guided radiographs had a significantly higher sensitivity for detecting tibial component loosening compared to standard radiographs (85.3% vs 74.8%, P = .02). Sensitivity in detecting femoral component loosening was poor overall and not improved by fluoroscopic enhancement compared to standard radiographs (58.8% vs 66.5%, P = .33). Fluoroscopically guided radiographs did not improve the specificity of detecting well-fixed implants in either tibial or femoral components nor affect the mean interobserver reliability over standard radiographs (kappa = 0.58 vs kappa = 0.60, P = .6).
Fluoroscopically assisted radiographs increased the sensitivity of detecting tibial component loosening over standard radiographs, but this clinical significance is unclear. Fluoroscopically guided radiographs may provide benefit in diagnosing aseptic loosening in select patients with painful TKAs.