Dynamic Movement Assessment and Functional Movement Screening for injury prediction: a systematic review

ABSTRACT Dynamic Movement AssessmentTM (DMATM) and Functional Movement ScreeningTM (FMSTM) are tools to predict the risk of musculoskeletal injuries in individuals who practice physical activities. This systematic review aimed to evaluate the association of DMATM and FMSTM with the risk of musculoskeletal injuries, in different physical activities, categorizing by analysis. A research without language or time filters was carried out in November 2016 in MEDLINE, Google Scholar, SciELO, SCOPUS, SPORTDiscus, CINAHL and BVS databases using the keywords: “injury prediction”, “injury risk”, “sensitivity”, “specificity”, “functional movement screening”, and “dynamic movement assessment”. Prospective studies that analyzed the association between DMATM and FMSTM with the risk of musculoskeletal injuries in physical activities were included. The data extracted from the studies were: participant’s profile, sample size, injury’s classification criteria, follow-up time, and the results presented, subdivided by the type of statistical analysis. The risk of bias was performed with Newcastle-Ottawa Scale for cohort studies. No study with DMATM was found. A total of 20 FMSTM studies analyzing one or more of the following indicators were included: diagnostic accuracy (PPV, NPV and AUC), odds ratios (OR) or relative risk (RR). FMSTM showed a sensitivity=12 to 99%; specificity=38 to 97%; PPV=25 to 91%; NPV=28 to 85%; AUC=0.42 to 0.68; OR=0.53 to 54.5; and RR=0.16-5.44. The FMSTM has proven to be a predictor of musculoskeletal injuries. However, due to methodological limitations, its indiscriminate usage should be avoided.


INTRODUCTION
Musculoskeletal injuries are one of the main causes of morbidity in individuals who practice physical exercises 1-3 . Thus, several screening methods have been developed aiming at classifying the risk of injury. In this context, functional tests based on subjective evaluations have been increasingly performed 4 to verify the movement patterns and dysfunctions associated with injuries of the trunk and lower limbs 5,6 . The subjectivity of these evaluations limits their reliability 7 . However, they are a low-cost alternative in large-scale evaluations and in case of absence of gold standards 8 .
To establish a risk classification tool for musculoskeletal injuries, Cook et al. 9 developed the Functional Movement Screening™ (FMS™). This method classifies the risk of injury in the presence of abnormal movement patterns by performing seven tests/ movements 9,10 . Each test can be evaluated from zero to three points and assess the interactions of kinetic chain mobility and stability needed to perform fundamental movement patterns. The score ranges from 0 to 21 points.
Initial studies have shown that soccer players with scores of 14 or less in the total score have a higher risk of injury 11 . Thus, this method has been used in preseasons of several modalities of sports to modify movement patterns that can cause injuries 10,12 . However, the efficacy of FMS ™ to predict injuries is controversial among authors [13][14][15] , likely justified by the different demands among sports 13 .
Years later, Nessler & Dunn developed the Dynamic Movement Assessment™ (DMA™) 5 . It consists of filming the individual performing six functional tests. The video analysis is performed using a twodimensional (2D) biomechanical analysis software. The 2D evaluation allows visualizing movement dysfunctions in the frontal plane, such as dynamic valgus 8 . Each of the six DMA™ tests is rated with a score ranging from zero (if the pain is related to the test) to three points. Each test has a major deviation and secondary deviations, which are observed. Failing to perform the test, the presence of three minor deviations, or a major deviation of greater magnitude characterizes a one-point score. The presence of two secondary deviations or a major deviation with intermediate magnitude promotes two points. Finally, individuals who perform the test without clinically important deviations are classified with three points 5 .
Due to the low cost and easy feasibility of the FMS™ and DMA™, their use to evaluate individuals who practice physical exercises in several groups is attractive. The main difference between the two methods is that DMA™ is based on functional tests with unilateral support (squatting and vertical jump), common in the sport gesture of several modalities 5 .
The fact that the incidence of injuries in people who practice sports 16,17 or occupational physical exercises 1,18,19 is high justifies this review. Getting to know a low-cost and easy-to-use test that measures fundamental movement dysfunctions, potentially predicting athletic injuries, may allow the development of preventive strategies that avoid the removal of functions involving physical exercises. Moreover, previous reviews evaluated only the diagnostic accuracy indicators of prospective studies of FMS™. Thus, the purpose of this systematic review is to evaluate the association of DMA™ and FMS™ with the risk of musculoskeletal injuries.

METHODOLOGY
This systematic review was registered in PROSPERO (CRD42017068014) and drafted based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement 20 .

Inclusion criteria
The studies included in this systematic review were prospective studies that used the FMS™ or DMA™ to classify the risk of musculoskeletal injuries in physical exercise practitioners of both sexes and without age limits. More detailed information about FMS™ and DMA™ are found in the studies of Cook et al., and Nessler & Dunn, respectively 5,9 .

Search strategy
A search was conducted in November 2016 in US National Library of Medicine (MEDLINE), Scientific Electronic Library Online (SciELO), Google Scholar, Virtual Health Library (VHL), CINAHL (EBSCOhost), SPORTDiscus and SCOPUS. The following keywords were used as descriptors of the Medical Subject Headings (MeSH): injury prediction, injury risk, functional movement screening, and dynamic movement assessment. The sentences used in this research were done with the Boolean operators AND (between the descriptors) and OR (between descriptor's synonyms). No date limits or language filters were applied.

Data collection process
The following data were extracted from the selected studies: profile of the participants, sample size, classification of musculoskeletal injuries, follow-up time and type of statistical analysis performed with its results.

Bias risk analysis
For bias risk analysis, the Newcastle-Ottawa scale was used 21 . The domains considered in this scale are: (1) selection (representativeness of the exposed cohort, selection of the unexposed cohort, evaluation of the exposure and confirmation that the result of interest was not present at the beginning of the study, (2) comparison of the cohort based on the study design or analysis (if results were adjusted for the main confounding factors and other variables) and (3) outcome (outcome assessment, sufficient follow-up time, and adequacy of cohort follow-up). Studies with less than five stars were classified as a "high risk of bias." In addition, studies were considered to have a "risk of uncertain bias" as they did not score in the "comparison" domain. The bias risk analysis was performed by only one evaluator.

Flow diagram
The total of studies per database and flow diagram of the studies are in Figure 1. Seven studies were manually located. None of the studies investigated the DMA™. Characteristics of the studies included are in Table 1. Statistical analysis and its result are shown in Table 2, and bias risk analysis are exposed in Table 3. Criteria used to define the injuries: I) the diagnosis of the injuries was performed by a healthcare professional; II) The registered injuries were related to the training or competition; and III) Only the musculoskeletal injuries with time-loss longer than 24 hours were considered lesions; III * =time-loss longer than 3 weeks; III ** =a severe injury promotes withdrawal from the training program; III # =time-loss longer than 7 days.

DISCUSSION
This review aimed to evaluate the association between FMS™ and DMA™ with the risk of musculoskeletal injuries. No studies with DMA™ were found, probably due to its recent development 5 . Based on the statistical analysis of most studies evaluated, FMS™ is associated with the risk of musculoskeletal injuries. Considering the cohort studies by Bushman et al. 22 and O'Connor et al. 27 , which had the lowest risk of bias, this association is strengthened (Table 3).
According to Table 2, the FMS™ showed sensitivity values ranging from 26 to 68%; specificity from 38 to 96%; PPV from 19 to 91%; NPV from 28 to 85%; and AUC from 0.42 to 0.68 (Table 2). Therefore, it is noticeable that the indicators of diagnostic accuracy are divergent between the 12 studies. OR values also ranged from 0.53 to 11.67, which corresponds, according to the literature, to absent and large effect sizes, respectively 41 . Only seven studies calculated the RR 20,27,34,35,40,42,43 , whose results were RR=1.86 (overuse injuries) 22 , RR=1.49 (traumatic injuries) 22 ; and RR=-0.5 40 to 2.73 39 (any injury). Thus, a low score in FMS™ is associated with higher injury risk, although this result is limited by the number of studies that calculate the RR and the high risk of bias in two of those studies 27,34 .
Considering the results of the studies with low risk of bias (Table 3), it is verified that FMS has a low sensitivity 20,22,40 , a good specificity 20,33 , and AUC values slightly above chance 20,22 . Three out of the four studies with low risk of bias used samples composed of soldiers 20,33,40 . These studies showed a higher score in the "selection" domain. This observation is a consequence of a greater representativeness of the samples in military courses, an adequate selection of the unexposed cohort (which is part of the same population) and the monitoring of the exhibition (based on the analysis of base records, such as military base records). At the same time, military groups are generally more homogeneous regarding various characteristics (age, level of fitness, volume of physical exercise, routine, etc.). Only six studies analyzed the influence of potential confounders [20][21][22]33,40,41 . Thus, all other studies have a "high" or "uncertain" risk of bias, once the influence of other risk factors on the result obtained was not reported. Most of the samples used had little representativeness, especially in studies with athletes. In addition, in many cases, attrition rates were neither quoted nor justified. In some cases, the absence of cases was not confirmed at the beginning of the studies, and it was not clear whether there was blinding of the participant and the professional responsible for the follow-up, which limits the interpretation of the results.
All the studies included in this systematic review had the same prospective design, in which the association between the score of FMS™ and the risk of injury were evaluated. Among 20 studies, 15 performed the diagnostic accuracy analysis, 12 calculated the OR and 7 the RR. The large variation between the results might have relation to several factors. First, samples consisted of athletes from different modalities or soldiers. In addition, the age of the individuals differed from one study to the other. Probably, FMS™ is not an appropriate assessment tool for every physical exercise practitioner. Second, the rating of injuries does not follow the same criteria in all studies. Some authors used the definition proposed by Hägglund et al. 42 , which defines a musculoskeletal injury when three criteria are related to injuries: association with athletic participation; necessity for health care; and time-loss with restrict participation for at least 24 hours. However, some authors included only severe injuries (with time-loss larger than three weeks) 11 or any injury 19,20,22,24,27,40,41 . Third, statistical analysis based on indicators of diagnostic accuracy or simple calculation of OR limits the interpretation 41 . Diagnostic accuracy indicators are found in studies evaluating the validity of an index test compared with a reference standard 43,44 . In injury prediction studies, considering the occurrence of injuries as a reference pattern may limit the interpretation of the results, since a high-risk individual may not suffer an injury, especially if he/she is not exposed to the risk factor. The use of OR evaluates the chance of a highrisk individual to develop injuries. However, it does not consider the injury incidence 41 . Therefore, the most appropriate calculation is the relative risk.
This systematic review was the first to evaluate the association of FMS™ and DMA™, categorizing by type of statistical analysis performed. However, the small number of studies evaluating the RR of FMS™ and the absence of studies with DMA™ were limitations. In future studies, the control of some biases is recommended. Most of the studies did not perform pairing of variables such as gender, age and other variables of interest, such as sport modalities 45 . In this case, we suggest using logistic regression analysis. Another critical point was the lack of confirmation of case of absences in the baseline, as well the non-blinding of the evaluators responsible for monitoring the sample. In theory, they should not know whether the participant belonged to the group exposed to the risk factor. Finally, the development of studies about the association of DMA™ with the risk of musculoskeletal injuries is suggested, since no studies with this method were found, which uses movements present in several sport gestures with two-dimensional analysis 5 .

CONCLUSION
From the studies of this systematic review, the conclusion was that movement dysfunction, evaluated by FMS™, may be associated with the risk of injury in people who practice physical exercises. No studies evaluating the association between the DMA™ score and the risk of injury were found. It is recommended that future studies carry out greater control of selection, comparison and outcome biases, and perform a meta-analysis.