Interventions designed to prevent anterior cruciate ligament injuries in adolescents and adults: a systematic review and meta-analysis.

BACKGROUND Anterior cruciate ligament (ACL) injuries are common, result in significant morbidity, and are expensive to repair surgically and to rehabilitate. Several randomized and observational studies have tested neuromuscular interventions as preventive measures for these injuries. PURPOSE To conduct a systematic review and meta-analysis of all known comparative studies for estimating and testing the effect of neuromuscular and educational interventions on the incidence of ACL injuries in adolescents and adults, both male and female. STUDY DESIGN Systematic review and meta-analysis. METHODS Several databases were used to identify eligible studies through July 4, 2011: MEDLINE, EMBASE, SPORTDiscus, Cumulative Index to Nursing and Allied Health Literature, the Cochrane Central Register of Controlled Trials, and Health Technology Assessment. Eligible studies were assessed for risk of bias, and meta-analyses were performed on the estimated intervention effect (log incidence rate ratio) using inverse-variance weighting, subgroup analysis, and random-effects meta-regression to estimate the overall (pooled) effect and explore heterogeneity of effect across studies (measured by I2 and tested with the Q statistic). RESULTS Eight cohort (observational) studies and 6 randomized trials were included, involving a total of approximately 27,000 participants. The random-effects meta-analysis yielded a pooled rate-ratio estimate of 0.485 (95% confidence interval [CI], 0.299-0.788; P = .003), indicating a lower ACL rate in the intervention groups, but there was appreciable heterogeneity of the estimated effect across studies (I2 = 64%; P = .001). In the meta-regressions, the estimated effect was stronger for studies that were not randomized, performed in the United States, conducted in soccer players, had a longer duration of follow-up (more than 1 season), and had more hours of training per week in the intervention group, better compliance, and no dropouts. Nevertheless, residual heterogeneity was still observed within subgroups of those variables (I2 > 50%; P < .10). CONCLUSION The authors found that various types of neuromuscular and educational interventions appear to reduce the incidence rate of ACL injuries by approximately 50%, but the estimated effect varied appreciably among studies and was not able to explain most of that variability. CLINICAL RELEVANCE Neuromuscular and educational interventions appear to reduce the incidence rate of ACL injuries by approximately 50%.