Does research actually support practice in professional football? Thoughts on the BJSM review.
Updated: Aug 23, 2021
Last month a research group published an article in the British Journal of Sports Medicine that had such an important message that I really wanted to write some comments about it.
In their 2014 study, McCall and colleagues surveyed the most popular practices and perceptions in premier league football teams from around the world in terms of injury risk factors, screening tests and preventative exercises. In the follow up study published last month (McCall et al, 2015) they reviewed the literature to assess if the published research actually supported these. The findings are important for us all to consider, whether you are from an applied or research perspective.
What was the justification for the research?
In work previously published by the research group (McCall et al, 2014) the current perceptions and practices of 44 international premier division football (soccer) regarding non-contact injuries were surveyed.
Top Five Responses (rank order)
Perceived Risk Factors
Functional movement screen
Nordic hamstrings AND gluteus activation (joint fifth place)
Previously no systematic review exists into injury prevention and professional football that assigns a specific level of evidence for the consideration of risk factors and/or use of specific screening tests and preventative exercises based on the quality of studies. This is important from two perspectives:
• Practitioners should be guided by research to be confident that they are implementing current based evidence-based practice
• Researchers should be guided on future research that will help impact applied practice
Therefore in the current study (McCall et al, 2015) the aim of the research was:
‘To systematically review the research literature for the aforementioned ‘Top 3’ risk factors, screening tests and preventative exercises and to provide a graded recommendation for their use and consideration in practice.’ (p1)
The authors carried out a systematic review of PubMed (MEDLINE) and SportDiscus databases in February 2014, and an identical search again during September 2014 in response to journal reviewer comments, adding PEDRO and Cochrane databases. The population inclusion criteria had to consist of ‘elite’ (defined as a player playing professionally in at least the top 3 divisions of any country), male, Association football (soccer), 18 years or older. Two external research experts in injury risk and prevention were also contacted to help reduce the risk of missing relevant articles. The paper shows their screening process from 792 original records to the final 14 articles included in the study.
Each of the articles included in the review were assessed for the strength of their evidence using a methodological quality score (%) from a validated checklist for the assessment of healthcare interventions (Downs and Black, 1998) and a level of evidence from 1++ to 4 (Harbour and Miller, 2001). The levels are based on the type of study; level 1 quality systematic reviews and meta-analyses, 2 cohort and case studies, 3 non-analytical studies and 4 expert opinion, with extra marks of ++, + and – for levels 1 and 2 depending on quality and bias. An overall level of evidence was assigned for each risk factor based on these scores plus the judgement of the research group, and graded recommendations were assigned for the screening tests and preventative exercises for use in the practical setting (A: Strong, B: Moderate, C: Weak or D: Insufficient evidence to make a specific recommendations).
The overall levels of evidence for each factor as presented in the abstract were as follows:
Perceived Risk Factors
Previous Injury 2++
Muscle imbalance Inconclusive
Functional movement screen D
Isokinetic dynamometry D
Eccentric exercise C
Hamstring eccentric D
The risk factors could not receive a graded recommendation as they are not a test or exercise you can prescribe and the overall level of evidence of the top 3 responses was mixed. Their level of evidence of ‘2++’ is the highest available for cohort studies and so this was probably the most positive area in terms of being linked to current research. No articles met the inclusion criteria for fatigue as a risk factor in professional footballers, demonstrating the complexity of the evaluation of fatigue itself. Fatigue is something we talk about day in, day out in the applied world but how much we actually ‘know’ about it scientifically is debatable. I believe the authors surmise this perfectly: ‘Before even contemplating to begin to quantify fatigue in football, it is imperative that consensus is achieved on the definition of fatigue and how to identify the different mechanisms in a variety of situations, for example, during match-play or as a result of fixture congestion’ (McCall et al, 2015, p5) Muscle imbalance was assigned as inconclusive due to the limited and contradictory research included in the review. As with the discussion on ‘fatigue’, the term ‘muscle-imbalance’ is a equivocal term that would benefit from being defined by a consensus. This also has implications for isokinetic testing as discussed later.
The graded recommendation for all of the top 3 screening tests was D (insufficient evidence to make a specific recommendation). The Functional Movement Screen (FMS) was found to be the most commonly used by teams in the previously published study (McCall et al, 2014). Although the authors list a number of limitations associated with the test, I believe many practitioners using FMS are aware of them (as of course we should be with any test or technology we are using in applied practice). This may also explain why many practitioners have adapted the test for their own in-house versions. Only one study in the literature met the inclusion criteria for using a questionnaire as a screening test. That study found coping with adversity was associated with injury and was given a level of evidence of 2++, but of course is a single article hence why a recommendation of D was given. A ‘D’ was also given for isokinetic muscle testing as the literature had mixed results with supporting its use with injury.
After assigning a graded recommendation of C for hamstring eccentric exercise specifically to prevent hamstring injury in professional footballers, even the authors note their surprise that the recommendation is weak given the considerable importance placed by practitioners in this area. As with the multidimensional prevention programmes in the practical setting, many studies involved eccentric exercise performed in conjunction with other exercise types. Eccentric exercises for other body parts, such as Achilles or patellar tendons, were assigned level 4 (expert opinion) and graded recommendation of D, with only 1 article included and no scientific evidence currently supporting their use with this population. Furthermore it seems this may be one of the most pertinent areas for future research given the widespread use in the applied environment and the potential for adverse effects, such as jumper’s knee. Finally, a graded recommendation of D was given to balance and proprioceptive prevention exercises for both knees and ankles. No adverse effects have been noted thus far for these exercises so while researchers should work to validate these exercises, practitioners have been advised they can keep them within their programmes.
The authors conclude: ‘Our systematic review shows that most of the perceptions and practices of practitioners are not supported by scientifically validated recommendations from research’. (McCall et al, 2015, p7)
While I personally would not expect to see complete alignment between the two worlds of Research and Applied Practice, as a practitioner I found it quite stark just how far apart the research and current perceptions and practices were. For all three of the screening tests and two of the top three preventative exercises to be given insufficient evidence to assign a specific recommendation, is an overwhelming message. As a practitioner I have always defined my role as a Sport Scientist to: “use research and technology to support the eye of the specialist”, however this systematic review would in fact suggest that the research does not support many of our practices. This of course is not to say that these tests and exercises are not valid in our environment but it is enough to state they are not yet supported by the research with our population.
On the plus side it was positive to see many of the included articles coming from Jan Ekstrand’s research group, using the UEFA Champions League auditing data. This reinforces the importance of practitioners participating in and providing due care to this line of work. Outside of the Champions League study data, many of the top leagues do not seem to appear in the articles so this is a key method for access to the top elite.
This links to the study limitation in defining ‘elite’ or ‘sub-elite’. This study used the top 3 divisions across any country in order to align with the previous study, however as discussed in the paper, an alternative definition may be those competing in major competitions or top countries of FIFA confederations. After all it is clear that the third division of one country will be far from the same playing level as the top division of another. However, clearly this would currently greatly limit the articles included in such a systematic review.
Ultimately I think this study reinforces the need for academic institutions and professional football clubs to continue their ongoing work together. I am about to embark on a University supported research project within my applied environment in professional football for the first time. Meanwhile many of my colleagues are also carrying out applied research. The more of this kind of work that can be published in the literature the better it is for closing the gap in applied practice and published evidence. A number of examples spring to mind already, such as the case studies published by Liverpool FC (Brukner et al, 2014; Milson et al, 2014), training and recovery analysis by the Manchester United group (Gaudino et al, 2014; Thorpe et al, 2014) and physical time-motion analysis from Newcastle FC (Ade et al, 2014), to name but a few.
One of the inherent differences between applied and academic research is that (generally) academic research deals with population trends and averages, whereas in the applied world we are most interested in the individuals. Outliers that get swallowed up by the mean statistics of published studies are often of most interest to us in the applied world; perhaps these outliners are the elite individuals we are dealing with. Some of the case study research previously mentioned is one way around this.
This is not a new discussion of course. In 2008, Dr David Bishop presented an Applied Research Model for the Sport Sciences (ARMSS) in response to anecdotal evidence that suggested sport science research then was not informing sport science practice. So it seems the discussion is just as pertinent today in 2015. Ultimately we need to try to improve the overlap between the two worlds of Academic and Applied research (see figure 2).
Without doubt there is always an ‘art’ to employing practices/technology/tests in the applied environment, blending academic research with practitioner and peer experience and the vital ingredient of common sense! But at the end of the day we are ‘Scientists’ so I hope in the future we can continue to align research with applied practice.
Ade JD, Harley JA and Bradley PS (2014) Physiological response, time-motion characteristics, and reproducibility of various speed-endurance drills in elite youth soccer players: small-sided games versus generic running. Int J Sports Physiol Perform 9(3): 471-9.
Bishop D (2008) An applied research model for the sport sciences. Sports Med 38(3): 253-63.
Brukner P, Nealon A, Morgan C, et al. (2014) Recurrent hamstring muscle injury: applying the limited evidence in the professional football setting with a seven-point programme. Br J Sports Med 48: 929-38.
Downs SH and Black N (1998) The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health 52: 377-84.
Gaudino P, Alberti G and Iaia FM (2014) Estimated metabolic and mechanical demands during different small-sided games in elite soccer players. Hum Movement Sci 36: 123-33.
Harbour R and Miller J (2001) A new system for grading recommendations in evidence based guidelines. Br Med J 323: 334-6.
McCall A, Carling C, Nedelec M, et al. (2014) Risk factors, testing and preventative strategies for non-contact injuries in professional football: current perceptions and practices for 44 teams from various premier leagues. Br J Sports Med 48: 1352-7.
McCall A, Carling C, Davison M, et al. (2015) Injury risk factors, screening tests and preventative strategies: a systematic review of the evidence that underpins the perceptions and practices of 44 football (soccer) teams from various premier leagues. Br J Sports Med Published Online First: doi:10.1136/bjsports-2014-094104
Milsom J, Barreira P, Burgess DJ, et al. (2014) Case Study: Muscle atrophy and hypertrophy in a premier league soccer player during rehabilitation from ACL injury. Int J Sport Nutr Exerc Metab. 24(5): 543-52.
Thorpe R, Strudwick A, Buchheit M, et al (2014) Monitoring recovery during the in-season competitive phase in elite soccer players. In: World Conference on Science and Soccer 4.0 Abstract Book, June 5-7 2014, Portland, Oregon, USA: 21.