Objective To systematically review and summarise the evidence for the effects of neuromuscular training compared with any other therapy (conventional training/sham) on knee proprioception following anterior cruciate ligament (ACL) injury.Design Systematic Review.

Data sources PubMed, CINAHL, SPORTDiscus, AMED, Scopus and Physical Education Index were searched from inception to February 2020.

Eligibility criteria Randomised controlled trials (RCTs) and controlled clinical trials investigating the effects of neuromuscular training on knee-specific proprioception tests following a unilateral ACL injury were included.

Data extraction and synthesis Two reviewers independently screened and extracted data and assessed risk of bias of the eligible studies using the Cochrane risk of bias 2 tool. Overall certainty in evidence was determined using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) tool.Results Of 2706 articles retrieved, only 9 RCTs, comprising 327 individuals with an ACL reconstruction (ACLR), met the inclusion criteria. Neuromuscular training interventions varied across studies: whole body vibration therapy, Nintendo-Wii-Fit training, balance training, sport-specific exercises, backward walking, etc. Outcome measures included joint position sense (JPS; n=7), thresholds to detect passive motion (TTDPM; n=3) or quadriceps force control (QFC; n=1). Overall, between-group mean differences indicated inconsistent findings with an increase or decrease of errors associated with JPS by ≤2°, TTDPM by ≤1.5° and QFC by ≤6 Nm in the ACLR knee following neuromuscular training. Owing to serious concerns with three or more GRADE domains (risk of bias, inconsistency, indirectness or imprecision associated with the findings) for each outcome of interest across studies, the certainty of evidence was very low.

Conclusions The heterogeneity of interventions, methodological limitations, inconsistency of effects (on JPS/TTDPM/QFC) preclude recommendation of one optimal neuromuscular training intervention for improving proprioception following ACL injury in clinical practice. There is a need for methodologically robust RCTs with homogenous populations with ACL injury (managed conservatively or with reconstruction), novel/well-designed neuromuscular training and valid proprioception assessments, which also seem to be lacking.

PROSPERO registration number CRD42018107349.