Poor Validity of Functional Performance Tests to Predict Knee Injury in Female Soccer Players With or Without Anterior Cruciate Ligament Reconstruction

Background: Various tests have been developed to evaluate athletes functional performance and for use as screening tools for injury prediction. Further validation of their accuracy to predict injury is needed. Purpose: To investigate the validity of predetermined cutoffs used to differentiate between high- and low-risk players in different functional performance tests to predict (1) anterior cruciate ligament (ACL) injury or (2) severe traumatic knee injury in a cohort of female soccer players with a primary unilateral ACL reconstruction and a cohort of knee-healthy players. Study Design: Cohort study; Level of evidence, 2. Methods: A total of 117 active female soccer players (mean age +/- SD, 20 +/- 2 years) an average of 19 +/- 9 months after ACL reconstruction and 119 knee-healthy players (age, 19 +/- 3 years) were prospectively followed up for 2 years for new knee injuries. At baseline, all players underwent tests to assess postural control (Star Excursion Balance Test), hop performance (single-leg hop for distance, side hop), and movement asymmetries in the lower limbs and trunk (drop vertical jump [DVJ], tuck jump). The predictive validity of the test cutoffs to identify players who would sustain an ACL injury or a severe traumatic knee injury (absence from soccer play, >28 days) was assessed. The risk ratio (RR), area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were calculated. Results: A total of 46 players (39%) with ACL reconstruction sustained 48 severe knee injuries, including 28 ACL ruptures. Of the knee-healthy players, 13 (11%) sustained 14 severe knee injuries, including 8 ACL ruptures. No association was found between the predetermined functional performance test cutoffs and the risk of a new ACL injury or severe knee injury in players with ACL reconstruction. In knee-healthy players, the only variable associated with future ACL injury was >= 6.5 cm knee valgus in the frontal plane (any knee) in the DVJ (RR, 4.93; 95% CI, 1.04-23.40; P = .045), but with only fair predictive validity (AUC, 0.7; sensitivity, 0.75; specificity, 0.65). Conclusion: In our cohorts of female soccer players, the validity of commonly used functional performance tests to predict new knee injuries was poor. Only knee valgus during the DVJ was associated with new ACL injuries in knee-healthy players, but with only fair predictive validity