Comparative Outcomes of Direct Versus Connector-Assisted Peripheral Nerve Repair

Background: Peripheral nerve injuries affect a significant proportion of patients with upper extremity trauma, with transections frequently requiring surgical intervention. While direct repair (DR) remains the historical standard, connector-assisted repair (CAR) has been proposed to improve functional outcomes by addressing limitations inherent to DR, such as fascicular misalignment and tension at the repair site. Objectives: The purpose of this systematic review is to evaluate and compare the clinical effectiveness and complication rates of DR versus CAR in upper extremity peripheral nerve injuries. Methods: A systematic search of the PubMed, Scopus, and Ovid MEDLINE databases was conducted for clinical studies published between January 1980 and August 2025 that reported sensory outcomes after DR or CAR for peripheral nerve injuries in the upper limb. Studies were included if sensory outcomes could be categorized using the Medical Research Council Classification (MRCC) scale. The primary outcome was the rate of meaningful sensory recovery (MR), defined as MRCC ≥ S3, with a secondary threshold of MRCC ≥ S3+. Secondary outcomes included postoperative neuroma formation, cold intolerance, pain scores, altered sensation, and revision rate. Statistical analysis was performed using two-sided Fisher exact tests and unpaired t-tests, with p 15 mm. Motor recovery was limited, with better values in DR. DASH scores averaged 13.2 (DR) and 18.2 (CAR), with follow-up of 26 and 23.8 months. Complications were more frequent in DR for cold intolerance, altered sensation, and pain, whereas neuromas, revisions, and fistulas were higher in CAR. Conclusions: Connector-assisted repair demonstrates better sensory recovery and less cold intolerance than DR in small-gap upper-extremity nerve injuries but with higher post-interventional risks and costs. DR remains effective for closely approximated nerves. Randomized trials are warranted, as current evidence is heterogeneous and mostly observational.