Purpose: Currently, there is little information available on the management and outcome of intra-operative periprosthetic humeral fractures during shoulder arthroplasty. The purpose of this study was to report on the incidence, management, and outcome of intra-operative periprosthetic humeral fractures.

Method: Between 1980 and 2002, forty-six intra-operative periprosthetic humeral fractures occurred during shoulder arthroplasty at our institution. Thirty-six fractures occurred during primary total shoulder arthroplasty, five during primary hemiarthroplasty and four during revision shoulder arthroplasty. Twenty-one fractures involved the greater tuberosity, 16 the humeral shaft, 6 were metaphyseal and 2 fractures involved the greater tuberosity and extended to the humeral shaft. All patients were followed for a minimum of two years (mean, 7.5 years). At final follow-up, the Simple Shoulder Test (SST) and ASES scores were calculated and preoperative, postoperative, and most recent radiographs were examined for fracture healing and implant integrity.

Results: The incidence of intra-operative humerus fractures at our institution was 1.5%. All fractures healed at mean of 13 weeks (range, 6 to 56 weeks). At a mean final follow-up of 90 months (range, 25 to 194 months), the mean forward elevation was 108° and the mean external rotation was 39°. The mean SST was 6 and the mean ASES score was 61. Four patients underwent revision shoulder surgery, two for shoulder instability, one for a massive rotator cuff tear, and one for another periprosthetic humerus fracture that occurred after a fall. Complications included transient nerve injuries in six patients and four fractures displaced postoperatively which were then managed nonoperatively to healing.

Conclusion: The data from this study suggest that although intra-operative fractures have a high rate of healing, there was a significant rate of associated complications including transient nerve injuries and fracture displacement. Patients with intra-operative fractures also appear to have less forward elevation at final follow-up when compared to patients who have not sustained an intra-operative humeral fracture.

Ulnar neuropathy presents as a complication in 5% to 10% of total elbow replacements, but subsequent ulnar neurolysis is rarely performed. Little information is available on the surgical management of persistent ulnar neuropathy after elbow replacement. We describe our experience with the surgical management of this problem.

Of 1607 total elbow replacements performed at our institution between January 1969 and December 2004, eight patients (0.5%) had a further operation for persistent or progressive ulnar neuropathy. At a mean follow-up of 9.2 years (3.1 to 21.7) six were clinically improved and satisfied with their outcome, although, only four had complete recovery. When transposition was performed on a previously untransposed nerve the rate of recovery was 75%, but this was reduced to 25% if the nerve had been transposed at the time of the replacement.

We have examined the relief from pain and the functional outcome in 18 patients who underwent resection arthroplasty of the shoulder as a salvage operation between 1988 and 2002. The indications included failed shoulder replacement in 17, with infection in 13, and chronic septic arthritis in one. The mean follow-up was 8.3 years (2.5 to 16.6). Two intra-operative fractures of the humerus occurred, both of which healed.

The level of pain was significantly decreased (t-test, p < 0.001) but five patients continued to have moderate to severe pain. The mean active elevation was 70° (0° to 150°) postoperatively and represented an improvement from 39° (0° to 140°) (t-test, p = 0.003), but internal and external rotation were hardly changed. The mean number of positive answers on the 12-question Simple Shoulder Test was 3.1 (0 to 12) but the shoulder was generally comfortable when the arm was positioned at rest by the side. The mean post-operative American Shoulder and Elbow Surgeon’s score was 36 (8 to 73).

Despite applying this procedure principally to failed shoulder replacements, the results were similar to those reported in the literature for patients after severe fracture-dislocation. Reduction of pain is possible in one half to two-thirds of patients. The outcome of this operation in providing relief from pain cannot be guaranteed, but the shoulder is usually comfortable at rest, albeit with profound functional limitations.