LAUSR

Objective: To determine the effectiveness of proprioceptive training on knee function and proprioception following anterior cruciate ligament reconstruction. Data sources: PubMed, EMBASE, The Cochrane Library, Ovid, EBMSCO-host, ScienceDirect, CNKI, VIP, WanFang Data and CBM were searched until 2nd October 2020. Review methods: Trials with proprioceptive training for patients with anterior cruciate ligament reconstruction were included. Study screening, data extraction, risk of bias and quality assessments were performed independently by two researchers. We performed a stratified analysis based on the quality of the study. Sensitivity analyses were performed if the heterogeneity was high. Results: Seventeen trials with 878 participants were included, and 12/17 with low quality. After stratified analysis, the pooled effect of high-quality studies showed significant improvement for proprioceptive training group in range of motion (P < 0.05, I2 = 0%), but no differences in Cincinnati knee rating system score (P > 0.05, I2 = 83%), hop test (P>0.05, I2 = 0%) and proprioception (P = 0.17, I2 = 77%) compared to conventional training group; while the pooled effect of low-quality studies showed significant differences in knee functional improvement (P < 0.00001, I2 = 69%), but no difference in proprioception (P > 0.05, I2 = 84%) between two groups. The pooled effect of all studies showed a significant benefit in both function and proprioception. Conclusion: The effects of proprioceptive training on knee functional and proprioceptive improvement after anterior cruciate ligament reconstruction is mixed. It is more likely that proprioceptive training in high-quality studies has few detectable effects and that low-quality studies show an effect because of an unconscious bias. A large well designed high-quality study needs to be undertaken in the future.