Introduction: Clinically evident neurological injury of the operated limb after total shoulder arthroplasty is not uncommon. The purpose of this prospective study was to determine the incidence of subclinical neurological lesions after reverse shoulder arthroplasty and anatomic shoulder arthroplasty (group control), and to correlate its occurrence to postoperative lengthening of the arm.

Method: We included all patients needing a total shoulder arthroplasty either anatomic or reversed. Each patient underwent a pre- and postoperative electromyography (EMG). This study focused on the clinical, radiological and EMG evaluation, with a measure of the lengthening of the arm in case of reversed shoulder arthroplasty according to a protocol previously validated.

Result: Between November 2007 and February 2009, we collected 41 patients (42 prostheses), including 23 anatomic (group 1) and 19 reverse (group 2) primary shoulder arthroplasties. The 2 groups were similar according to mean age, comorbidity, male/female ratio and nerve conduction abnormalities on EMG performed on an average of 10 days before surgery. Control EMG realized at an average of 3.6 weeks postoperatively showed in group 1, a plexus lesion due to an intra-operative complication. In group 2, we noticed 9 recent neurological damages (45% of cases) involving mainly the axillary nerve; 8 were rapidly regressive. The incidence of recent injury was significantly more frequent in group 2 (p=0.003) with a risk 10.4 times higher (95% CI 1.4, 74.8). Mean lengthening of the arm after a reverse was 3.1 cm ± 1.8 (range 0.2 to 5.9) compared to preoperative measurement and 2.4 cm ± 2.1 (range −0.5 to 5.8) compared with the normal contra-lateral side.

Discussion: The occurrence of peripheral neurological lesion following a reverse shoulder arthroplasty is common but usually transient. These lesions may cause postoperative pain, alter rehabilitation and can theoretically induce prosthetic instability. Lengthening of the arm is considered as one of the major factors responsible for this neurologic damage. Indeed, surgical dissection, compression phenomena by use of retractors or presence of hematoma, vascular injury, mobilization of the upper limb and possibly interscalene block are similar for the two types of prosthesis. Arm lengthening is thus a compromise between necessary retensionning of the deltoid for good mobility and instability avoidance, and lengthening which may be responsible for neurological lesions, acromial fractures and permanent arm abduction.