My reply to a comment on my professional facebook page.
I welcome your post, but since this is my page, always open to robust discussion, I reserve the right to discuss your points further.
If you start a comment with "only problem is..." you create an impression of making a counter point to a solution. You can see why I pulled you up on your point, since I was not posting about a solution, just a definition of terms and place of those terms in development of programming.
Secondly, if you continue "only problem is.." with "all", you immediately generalise. That's fraught with opportunity for exposure as being a statement of weakness.
Third, if you highlight movement screen, we have a different discussion point that is welcomed as a topic for conversation. It seems we may have a disconnect between what is understood by screen, or movement. As it goes, the current understanding I have, and thus use this definition, for movement, is mobility of segments with stability and control of the body segments. The current definition I have of a screen, and thus use, is akin to a filter, to catch something of interest. Ideally, a screen that filtering movement should identify problems with mobility of segments or control of segments. Thus, a movement screen should have criteria that indicates when mobility, stability or motor control is problematic. Those criteria have to be articulated, thus they have to be discussed, applied, critiqued, refined and eventually set by individuals of expertise, with consensus. Thus, they will always be messy to some extent, but should have been set through honest process. An honest mess.
That a movement screen is an unproven entity requires clarification. What would you have a movement screen prove? In light of the above definition of a movement screen, it should reveal problems with patterns of movement that are contributed to by mobility and/or stability, and/or control of mobility. I would contend that such screens exist, and to varying extents do they indeed reveal, consistently between raters, as written about in at least 18 papers of reliability, that problems can be rated on an ordinal scale.
Re Professor Roald Bahrs comment that movement screens (again a generalisation) may or may not measure something important has two factors that spring to mind.
First, I put it to you that a screen that reveals different ratings of movement patterns, requiring and/or revealing pain with movement, limitations and asymmetries in mobility, stability and control of mobility, is indeed important. Six of the top modifiable risk factors for future injury indeed include limitations and asymmetries in mobility, stability and control of movement. I’ve written about these, with references to 33 publications (non exhaustive) in my article Bulletproofing the volleyball knee.
Second, it may be contrary to earlier statements of Prof Bahr’s, from his paper No injuries, but plenty of pain? On the methodology for recording overuse symptoms in sports. Br J Sports Med, 43, 966-72.
In this paper, Professor Bahr wrote, “The following recommendations are made: (1) studies should be prospective, with continuous or serial measurements of symptoms; (2) valid and sensitive scoring instruments need to be developed to measure pain and other relevant symptoms; (3) prevalence and not incidence should be used to report injury risk; (4) severity should be measured based on functional level and not time loss from sports. In conclusion, new approaches are needed to develop more
appropriate methodology to quantify overuse injuries in studies.” I reference this article because Prof Bahr specifically recommends the reporting of pain, functional level (thus assuming there are different levels of function that should be measured on a scale of different levels), and injury risk. I further bring this up because the topic of injury risk is important to discuss. That a risk of injury exists does NOT mean it predicts injury. Injury prediction is very difficult due to the multifactorial nature. For example, if the risk of injury in any individual is 16% (an approximate figure for the risk of hamstring strain in AFL footballers, for example, since approximately 1 in 6 players at an AFL club will suffer a hamstring injury each season), then there are known factors that increase the likelihood of any one athlete being one of those six. They include age, race, previous history – all non-modifiable – and other factors that are modifiable, such as Thomas test knee angle, hamstring strength, ankle movement asymmetry, low back pain, not to mention variations in training load and speed of the game. An individual with all of those risk factors may not manifest as an injury, thus indicating that the presence of risk factors does not predict injury.
Professor Bahr comments on Mick Drew, Jill Cook and Craig Purdam’s paper. He repeats that “The key question asked is whether it is possible to use screening tests to identify who is at risk for a sports injury-in order to address the deficit through a targeted intervention programme.” The next line in his abstract has errors – “The paper demonstrates that to validate a screening test to predict and prevent sports injuries.” Immediately we see the error, a screening test is valid if it reveals risk factors, not if it predicts injury – a very important distinction. Screening tests do NOT predict injury, they reveal risk factors. Professor Bahr’s summary that “Unfortunately, there is currently no example of a screening test for sports injuries with adequate test properties” is a summary that has two errors within it – it is based on the error – screening tests are not to predict sports injuries, but to reveal risk factors to those whose responsibility it is to know about them and modify within athletic preparation programs. This summary statement is also false in its contention that there are no screens examples with adequate test properties – his definition of the test properties are that a) a prospective study should reveal a link between screening test and risk, and b) the test properties are examined in relevant populations. To reveal that statement as not completely true, Mike Lehr and colleagues prospectively studied military and athletic populations, with multiple screening tests that revealed risk factors for injury associated with subsequent increased prevalence of injury in those with more risk factors.
Professor Bahrs statement that “such tests are unlikely to be able to predict injury with sufficient accuracy” is somewhat correct and somewhat incorrect. Some studies, such as Mike Lehr’s, has indicated more ability to predict injury using screening tests. Secondly, remember that screenings aren’t designed to predict injury, but to reveal risk factors – not the same thing. That there is no pure correlation between risk factors and prediction is discussed above, due to the multifactorial nature of intrinsic factors, extrinsic factors and training load. Movement screens do not reveal issues with extrinsic factors and training load – massively important factors in injury risk.
Professor Bahrs final comment, related to Drew et al’s paper, revolves around “The final step needed is to document that an intervention programme targeting athletes identified as being at high risk through a screening programme is more beneficial than the same intervention programme given to all athletes.” This statement is in error – why would an intervention program, to address individual risk factors, be beneficial to an athlete who doesn’t have those risk factors? It wouldn’t. That’s ludicrous. An athlete with asymmetrical ankle dorsiflexion, a known risk factor for ankle sprains, who is provided with an intervention to normalise the range of motion, will have a reduction in their risk factor. The same intervention to an individual who does not have an asymmetry will not see a reduction in their risk… why? Because there has been no change to the athlete because there was no need nor room for change. This is a flawed step in the requirements.
As for variability in movement – I guess you are suggesting that ordinal scales which rate movement to reveal limitations and asymmetries against a consensus of criteria, thus non-linear scales, do not consider variation as important. Of course variation in movement occurs, thus it is important. Variation occurs because of context-specific requirements such as position, pattern and power. Variations in movement can still reveal risk factors. The two are not mutually exclusive. An individual with limited ankle dorsiflexion who squats will vary their movement to flex more at the hip and spine, or will lift their heel. That’s variation. Is it require to complete the squat? Yes. Does the varied pattern reveal possible limitations in mobility and control of mobility? Yes. Are those revelations of risk factors? Yes. Ankle dorsiflexion restrictions is a known risk factor. Spine flexion movement under load is a risk factor. The requirement for variation for movement is not an argument against a movement screen – it is a component of movement that is necessary to indicate whether such variations may reveal risk factors – that is why we look – to filter movement to see if there are risk factors.