Accurate risk assessment is crucial in preventing sports-related injuries. Physical therapists, strength and conditioning coaches, and other practitioners utilized the Functional Movement Screen (FMS) to evaluate athletes' injury risk. However, its predictive effect was unstable and results had shown inconsistency across various studies To summarise the reliability of FMS in predicting sports injuries as currently documented in the scientific literature, elucidate the reasons behind inconsistent findings, and highlight the issues associated with FMS use in injury prediction. The reliability and feasibility of comprehensive score, single sub-tests, number of single risk indicators, combinations of risk indicators and sub-tests, and integration of risk indicators/sub-tests with other exercise tests in predicting sports injuries are discussed separately The identified limitations in FMS-based injury prediction stem from the misuse of statistical methods, lack of internal consistency in the evaluation system, and ambiguity in injury definitions. Therefore, future efforts should focus on integrating FMS sub-tests or risk indicators with additional assessments, employing machine learning techniques to devise predictive models tailored for different types of sports injuries in different sports events and different populations, ultimately enhancing the capability to predict the risk of sports injuries with greater accuracy. • 1.Reliance on FMS comprehensive score for injury prediction presents significant methodological limitations, with current evidence suggesting limited clinical utility for standalone forecasting. • 2.While FMS effectively identifies asymmetries, movement dysfunctions, and provocative pain, its application for injury risk assessment should be refined rather than discarded. Specifically, individual sub-tests and risk indicators should be deconstructed and integrated with other specialized assessments using machine learning to enhance predictive efficiency. • 3.Inaccurate injury predictions stem primarily from the inherent limitations of comprehensive score, misapplied statistical methodologies, and vague impairment definitions.
Dong et al. (Wed,) studied this question.