Routine immediate postoperative radiographs rarely identify unknown complications after shoulder arthroplasty.

Background: Intraoperative complications after shoulder arthroplasty (SA) are uncommon, yet surgeons continue to obtain immediate postoperative radiographs despite prior literature questioning the efficacy of these images. There is a paucity of literature describing the role of immediate postoperative radiographs in revision SA. This study aimed to evaluate the utility of immediate postoperative radiographs in identifying intraoperative complications leading to a change in care after primary or revision SA. We hypothesized that routine postoperative ("survey") radiographs would not provide a high diagnostic yield in identifying early complications requiring a change in postoperative management.

Methods: A retrospective review of 4063 SAs (1298 primary anatomic total shoulder arthroplasties [aTSAs], 2162 primary reverse shoulder arthroplasties [RSAs], 129 revision aTSAs, and 474 revision RSAs) was performed from a multi-institutional arthroplasty database. All shoulders were evaluated with a single immediate postoperative (survey) radiograph that was read by a musculoskeletal radiologist and by the treating orthopedic surgery team. The radiology reports of all 4063 immediate postoperative radiographs were reviewed to determine whether complications had been identified. For patients in whom complications were identified, the medical record was then reviewed to determine whether the survey radiograph resulted in a return to the operating room or a change in management between the surgical procedure and the first postoperative clinic visit with radiographic evaluation.

Results: Complications were reported by a radiologist on the basis of immediate postoperative radiographs in 10 of 4063 shoulders (0.2%, or 1 of every 500 arthroplasties). Complications were reported after 4 primary RSAs, 4 revision aTSAs, and 2 revision RSAs; no complication was reported after any of the primary aTSAs. Of the 10 complications, only 3 were unknown to the treating surgeon following the operation and none altered immediate postoperative management. Early complications were reported at a rate of 1% after revision arthroplasty vs. 0.1% after primary arthroplasty (P = .001). Additionally, revision aTSA was found to have a higher rate of complications reported on radiographs, with a rate of 3.1%, compared with revision RSA, at 0.4% (P = .021).

Conclusion: The radiology reports of routine immediate postoperative radiographs rarely identified postoperative complications (0.2%). In all cases, postoperative management remained unaltered until the time of the first postoperative visit. The value and medicolegal implications of immediate postoperative radiographs may inform the decision to abandon obtaining these studies routinely.