Methods: The following electronic databases were assessed for relevant articles by two independent reviewers: Medline, Cochrane, Embase and CINAHL. Articles were selected using specific inclusion criteria. Methodological quality of observational and non-randomized studies was assessed using the MINORs scale, and a predetermined score 10 or greater was considered to indicate a high quality study. We independently extracted data for clinical outcomes from the relevant articles: range of motion, strength, quality of life, cost and complications.
Results: In total, 270 articles were identified. Fifty-eight articles were assessed as relevant. Our kappa agreement was 0.96. Fifty-three of these 58 studies were retrospective and 5 were prospective. Only 7 of the non-randomized articles were deemed as high quality using the MINORs scale. The most commonly measured outcome was strength via MRC scale (44 of 58 articles), followed by range of motion (18 of 58). Other reported factors included: time to follow-up (50 of 58 articles), injury-to-operation time interval (44 of 58), surgical complications (30 of 58), dominant handedness/side of injury (7 of 58), and quality of life (3 of 58). No studies mentioned cost. It was not possible to accurately pool the data because investigators used many varying modalities of outcome measurement.
Conclusions: Based on the current evidence, the claim of superiority of nerve transfers vs. tendon transfers remains unsubstantiated. To better determine the overall superiority of either tendon transfers or nerve transfers, more methodologically sound randomized controlled trials are needed.