Nerve alerts and peripheral nerve injury in patients undergoing shoulder arthroplasty with intraoperative neuromonitoring: a systematic review

Abstract

BackgroundIntraoperative neuromonitoring (IONM) during shoulder arthroplasty can be useful for detecting and preventing nerve deficits. The purpose of this investigation was to systematically review the literature for studies that used IONM to investigate the incidence of potential nerve injury during and after shoulder arthroplasty. MethodsA systematic review was performed using Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. PubMed, Scopus, and Cochrane Central Register of Controlled Trials were searched for level I-IV studies that reported on the incidence of nerve alerts and peripheral nerve injuries following shoulder arthroplasty with IONM. Data recorded included study demographics, IONM parameters, nerve alerts, and postoperative nerve injuries. Study methodological quality was analyzed using the Methodological Index for Non-Randomized Studies score. Forest plots of proportions were generated for nerve alerts and nerve injury. A P value of 0.05 was set as significant. ResultsThe initial search yielded 1556 articles (414 duplicates, 1142 screened, and 31 full-text review). Nine articles (699 patients, three somatosensory evoked potential (SSEP) and six transcranial electric motor evoked potential (MEP) studies; 53.6% females; mean age range, 58-74.8 years; mean Methodological Index for Non-Randomized Studies score, 15.4 ± 4.6) were included and analyzed. Five studies set the threshold for a nerve alert to be a ≥50% reduction in amplitude, whereas four studies set this threshold to be a ≥80% reduction in amplitude. Of the procedures performed, 44.9% were anatomic total shoulder arthroplasties (aTSAs), 42.1% were reverse total shoulder arthroplasties (rTSAs), 3.7% were hemiarthroplasties, and 14% were revision arthroplasties. In all but one study that utilized MEP for IONM, axillary nerve alerts were the most common (39%). The humeral and glenoid preparation/implant stages accounted for the highest percentage of nerve alerts in all studies. Overall, 47% of patients had at least one intraoperative nerve alert. A significantly higher percentage of patients monitored with MEP had a nerve alert compared to those monitored with SSEP only (56% vs. 39%; P < .001). The overall pooled percentage of patients who had evidence of a clinical nerve injury postoperatively was 5% [2%-13%]. All but two clinical neurological deficits resolved within six months, and one resolved within one year. ConclusionIONM during shoulder arthroplasty identifies and may help prevent nerve injury. IONM with MEPs detected significantly more nerve alerts than that with SSEPs. The axillary nerve was the most common peripheral nerve affected. Humeral and glenoid preparation/implant stages were responsible for most alerts. The advantages of IONM must be balanced with the time, cost, and intraoperative clinical ambiguity of responding to nerve alerts.