TY - JOUR T1 - Study of the measurement and predictive validity of the Functional Movement Screen JF - BMJ Open Sport & Exercise Medicine JO - BMJ OPEN SP EX MED DO - 10.1136/bmjsem-2018-000357 VL - 4 IS - 1 SP - e000357 AU - Fraser Philp AU - Dimitra Blana AU - Edward K Chadwick AU - Caroline Stewart AU - Claire Stapleton AU - Kim Major AU - Anand D Pandyan Y1 - 2018/05/01 UR - http://bmjopensem.bmj.com/content/4/1/e000357.abstract N2 - Objective The aim of the study was to evaluate the reported measurement capabilities and predictive validity of the Functional Movement Screen (FMS) for injury.Methods This was a prospective observational longitudinal study of 24 male footballers from a single team in England, alongside analysis of an existing database over one season (September 2015–May 2016). A preseason FMS was carried out with scores recorded by an experienced assessor and derived, retrospectively, from the three-dimensional movement data that were simultaneously captured. The assessor scores were compared with the photogrammetric system to determine measurement validity, and predictive validity was quantified by assessing sensitivity and specificity (cut-off score of 14).Results The real-time assessor score matched the photogrammetric score awarded for one of the participants, was higher than the photogrammetric system for 22 participants and was lower than the photogrammetric system in 1 participant. There was no discernible relationship between FMS scores and the competencies required to be met as per the rules articulated for the allocation of a score. A higher number of total injuries were associated with higher FMS scores, whether determined through real-time assessment or codification of kinematic variables. Additionally, neither method of score determination was able to prospectively identify players at risk of serious injury.Conclusion The FMS does not demonstrate the properties essential to be considered as a measurement scale and has neither measurement nor predictive validity. A possible reason for these observations could be the complexity in the instructions associated with the scale. Further work on eliminating redundancies and improving the measurement properties is recommended. ER -