Fluoroscopically positioned radiographs of total shoulder arthroplasty

Between January 1990 and January 1991, 87 patients (100 shoulder arthroplasties) were radiographically examined with both standard and fluoroscopically positioned views. The sets of radiographs were compared and assessed for acceptability (the ability to interpret the interfaces), presence or absence of radiolucent zones, and differences between size and location of radiolucent zones. Fifty-six of the standard radiographs were unacceptable because of obliquity of the x-ray beam protection; only three localized views were unacceptable (p < 0.0001). Nineteen of the unacceptable standard views had radiolucent zones detected on the comparable fluoroscopic views. Forty-four radiographic sets were acceptable for evaluation. Of these, 27 sets had no differences; standard views showed the interface changes better in one, and localized views were better in 76 (p < 0.001). Thus in 68 of 100 total shoulder arthroplasties evaluated, clinical information that was not seen on standard radiographs was identified on fluoroscopic views. We conclude that fluoroscopic positioning is superior to standard radiographs in detecting implant-bone interface changes.