- On June 9, 2017
- athlete monitoring, athlete monitoring, injury prediction, injury prediction, smartabase, smartabase, training load, training load
See below, how one of our AFL clients have used SMARTABASE to facilitate quality research that translates to the field.
What question were we trying to answer?
- Which of our measures, collected weekly, has the greatest association with non-contact injury in-season? (workload, subjective wellness, musculoskeletal screening)
- Does a combination of these measures allow us to better detect non-contact injury?
Why did we do it?
An abundance of athlete data is collected in elite team sport settings. Streamlining this data to flag those players at risk can often be a lengthy task. By determining a hierarchy of risk factors, weekly risk assessment can be fast-tracked and presented to the coach in a simple and understandable manner.
Furthermore, with a valid and reliable database spanning the last four years, we were able to combine these measures to determine a multivariate model of injury risk, while accounting for moderating and interaction effects of all our key measures.
How did SMARTABASE facilitate this research?
Working with large datasets can be a tedious process. For this type of research, it is not uncommon to spend 90% of your time preparing the data for analysis — this is where SMARTABASE excelled and saved a lot of time:
- We were able to efficiently upload daily workload data which automatically calculated our 1-,2-,3-,4- weekly cumulative workloads and the acute:chronic workload ratio (ACWR) for each player.
- Furthermore, we had players enter their subjective wellness ratings straight into the SMARTABASE mobile application, saving any double handling of the data.
- Another unique feature was the ability to compare to a player’s historical data. For wellness ratings and musculoskeletal results, we were able to calculate individualised rolling averages for comparison and determine if a player had significantly deviated from their norm. The reporting tool allowed us to quickly view those players that elicited a “red flag” (1 standard deviation drop).
- Lastly, the ability to combine multiple data sources (workload, injury, screening, wellness) into one central form via linked calculations was crucial in reducing the data preparation time.
What did we find?
- A low four-week chronic load as of the Monday was associated with the greatest risk of injury in the subsequent 7-days. Being under-loaded and/or ill-prepared, may be mediator for non-contact injury.
- When we coupled four-week chronic load with the ACWR we saw a different relationship. High chronic workloads were protective while low chronic load coupled with both low and very high ACWR increased injury risk.
- As playing experience increased, so did injury risk. Our players with >9 years were associated with the greatest risk.
- Subjective wellness questions were associated with elevated risk in the subsequent seven-days, suggesting a player’s response plays a role in the multifactorial nature of injury risk.
- Our multivariate model performed significantly better than univariate models at detecting non-contact injury in the subsequent week.
How is this research practically applied?
- To mitigate risk, sports medical/ science staff use a SMARTABASE training drill database to derive individualised workload averages. This allows us to forecast loads and implement specific injury prevention strategies to ensure players are maintaining the required volume for competitive demands in-season.
- Measures are considered in combination as injury is multifactorial and we know some workload-injury relationships can be moderated (i.e. chronic workload and the ACWR)
- Players with greater playing experience are carefully managed as we know they’re at greater risk. We’re more likely to intervene with a training intervention if a load spike occurs for these players.
- Our subjective wellness ratings often act as a cue for further conversation. This allows our coaches and physios to follow up on any unreported soreness or screening flags which may need to be addressed before subsequent training planning.
If you are interested in hearing more about the SMARTABASE system, do not hesitate to contact us at email@example.com.
Title: Multivariate modelling of subjective and objective monitoring data improve the detection of non-contact injury risk in elite Australian footballers
Authors: Marcus J Colby a,b, Brian Dawson a,b, Peter Peeling a,c, Jarryd Heasman b, Brent Rogalski b, Michael K Drew d,e ,Jordan Stares a,b , Hassane Zouhal f and Leanne Lester a
Institution and affiliations:
a School of Sport Science, Exercise and Health, The University of Western Australia, Australia
b West Coast Eagles Football Club, Australia
c Western Australian Institute of Sport, Mt Claremont, Australia.
d Australian Institute of Sport, Canberra, Australia.
e Australian Centre for Research into Injury in Sport and its Prevention (ACRISP), Federation University Australia, Ballarat, Australia.
f Movement, Sport and Health Sciences laboratory (M2S). UFR-APS, University of Rennes 2 – ENS Rennes, Avenue Charles Tillon, CS 24414, 35044 Rennes Cedex, France.
By Marcus Colby, PhD Candidate at The University of Western Australia