Previous studies report the neurological complication rate for shoulder arthroplasty to be 4.3% to 5.0%, However, these studies were limited to total shoulder arthroplasty (TSA) and did not include hemiarthroplasty (HA) or reverse prosthesis arthroplasty (RPA). Our hypotheses were that the neurological complication incidence after shoulder arthroplasty would vary by type of procedure performed and that the overall incidence would be higher than previously reported in the literature.

We retrospectively reviewed the charts of 307 consecutive patients who had a total of 349 SA by the same surgeon between June 1995 and August 2007. Only patients with over six months follow up were included. The charts were reviewed for any sensory or motor disturbance postoperatively. Those who had EMG confirmation of nerve injury (NI) were placed into the surgical complication group, with a second group composed of patients with neurological symptoms (NS) who did not require electromyography (Dr Ji or Matt---how many in the NI group did not have EMG?). These two groups were statistically compared to those patients without neurological injury using standard statistics software. There were 113 HA, 191 TSA and 45 RPA with over 6 month follow up, and there were 10 (10/349; 2.9%)neurological injuries (NI) There was no significant difference in the incidence between the groups (HA: N=3/113, 2.7%; TSA: N=5/191, 2.6%; RPA:N=2/45, 4.4%). There were an additional 34 neurological symptoms (NS) after shoulder arthroplasy, and if included with the NI then the total rate of neurological complaints after shoulder arthroplasty was 12.6% (44/349). If the NI and NS are combined, multivariate analysis showed that there was a statistically significant association between the development of neurological symptoms and revision surgery.

The rate of neurological complications after shoulder arthroplasty was independent of the type of procedure. The incidence of neurological complaints after shoulder arthroplasty is higher than previously reported.