Incidence of peripheral nerve injury in revision total shoulder arthroplasty: an intraoperative nerve monitoring study.

Background: The incidence of nerve injuries in revision total shoulder arthroplasty (TSA) is not well defined in the literature and may be higher than that in primary procedures, with 1 study reporting a complication rate of 50% for shoulder revisions. Given that continuous intraoperative nerve monitoring (IONM) can be an effective tool in diagnosing evolving neurologic dysfunction and preventing postoperative injuries, the purpose of this study was to report on IONM data and nerve injury rates in a series of revision TSAs.

Methods: A retrospective cohort review of consecutive patients who underwent revision TSA was performed from January 2016 to March 2020. Indications for revision included infection (n = 7); failed total arthroplasty and hemiarthroplasty secondary to pain, dysfunction, and/or loose components (n = 36); and periprosthetic fracture (n = 1). Of the shoulders, 32 underwent revision to a reverse TSA, 6 underwent revision to an anatomic TSA, and 6 underwent spacer placement. IONM data included transcranial electrical motor evoked potentials (MEPs), somatosensory evoked potentials, and free-run electromyography. The motor alert threshold was set at ≥80% signal attenuation in any peripheral nerve. Patients were screened for neurologic deficits immediately following surgery, prior to administration of an interscalene nerve block, and during the first 2 postoperative visits. Additional data collection included surgical indication, sex, laterality, age at surgery, procedure performed, body mass index, history of tobacco use, Charlson Comorbidity Index, medical history, and preoperative range of motion.

Results: A total of 44 shoulders in 38 patients were included, with a mean age of 63.2 years (standard deviation, 13.0 years). Of the procedures, 22.4% (n = 10) had an MEP alert, with 8 isolated to a single nerve (7 axillary and 1 radial) and 1 isolated to the axillary and musculocutaneous nerves. Only 1 patient experienced a major brachial plexus alert involving axillary, musculocutaneous, radial, ulnar, and median nerve MEP alerts, as well as ulnar and median nerve somatosensory evoked potential alerts. Age, sex, body mass index, Charlson Comorbidity Index, and preoperative range of motion were not found to be significantly different between cases in which an MEP alert occurred and cases with no MEP alerts. In the postoperative period, no minor or major nerve injuries were found whereas distal peripheral neuropathy developed in 4 patients (9.1%).

Conclusion: Among 44 surgical procedures, no patients (0%) had a major or minor nerve injury postoperatively and 4 patients (9.1%) complained of distal peripheral neuropathy postoperatively. In this study, we have shown that through the use of IONM, the rate of minor and major nerve injuries can be minimized in revision shoulder arthroplasty.