The incidence and decompression rates of median and ulnar neuropathies following shoulder surgery

Upper extremity peripheral neuropathy is a known, but uncommon complication that can occur after shoulder surgery. The incidence rate is well documented, and most of these cases historically have been treated conservatively. However, we hypothesize peripheral compression neuropathy requires a much higher need for surgical decompression than originally reported. The purpose of this study was to evaluate the incidence, decompression rates, and prognostic factors for developing median and ulnar neuropathies following shoulder surgery. A retrospective chart review was performed examining patients who underwent open and arthroscopic shoulder surgery from a multisurgeon, single-institution database. Perioperative data and functional outcomes were recorded. Symptom resolution was assessed with both conservative and surgical management of compression peripheral neuropathy. Further analysis was made to compare open and arthroscopic procedures, the type of neuropathy developed, and electromyographic (EMG) severity. The incidence rates of compression peripheral neuropathy following open arthroplasty and arthroscopic procedures was 1.80% (31/1722) and 0.54% (44/8150), respectively. 73.33% (55/75) of patients developed ipsilateral disease, while 20.00% (15/75) of patients had bilateral disease. Amongst the 75 included patients, there were 99 cases of neuropathy. Carpal tunnel syndrome was more common than cubital tunnel syndrome, comprising 61.61% (61/99) cases of neuropathy. 12.00% (9/75) of patients developed both carpal tunnel syndrome and cubital tunnel syndrome. Four patients were lost to follow-up. Decompression surgery was performed for 36.84% (35/95) cases of neuropathy with >90% obtaining symptom resolution with surgery. 63.16% (60/95) cases of neuropathy were managed conservatively, 71.67% (43/60) of which had persistent symptoms. When comparing arthroscopic versus open procedures, patients who underwent open procedures were significantly older (68.62 vs. 49.78 years, P