This week’s EXSS Impact Post is by Dr. Zack Kerr, Assistant Professor in Exercise and Sport Science. Dr. Kerr is an epidemiologist and explores different reporting methods of injury epidemiology data to improve public dissemination and understanding.
Why did you do this study?
Over the past decade, I have immersed myself in the world of sports injury surveillance. The findings are of great importance because they are the basis for many data-driven decisions regarding the rules and safety in sports organizations such as the National Collegiate Athletic Association (NCAA) and the National Federation of State High School Associations (NFHS).
However, at times, I found it difficult to discuss findings to stakeholders of sports organizations, such as parents, coaches, and administrators, because we in academia like to rely on the injury rate to measure injury incidence. The injury rate is defined as:
Injury rates are preferred because they account for all injury events in the numerator, regardless of whether or not these were sustained by the same athletes. They also account for variation in the amount of exposure time via the denominator; thus, an athlete that plays more across a season provides more exposure time. But for many parents, what matters most to them is knowing the risk; in other words, what is the probability that their child will get injured within a specific timeframe (e.g. one season).
As a result, using data from the NCAA Injury Surveillance Program, I worked with a team of epidemiologists and athletic trainers to examine a variety of methods of measuring injury incidence. This team included:
- Karen G. Roos, California State University – Long Beach
- Aristarque Djoko, Datalys Center for Sports Injury Research and Prevention
- Sara L. Dalton, Datalys Center for Sports Injury Research and Prevention
- Steve P. Broglio, University of Michigan
- Stephen W. Marshall, University of North Carolina at Chapel Hill
- Thomas P. Dompier, Datalys Center for Sports Injury Research and Prevention
Given the interest, we opted to examine these measures with concussion.
What did you do and what did you find in this study?
We used concussion data from the NCAA Injury Surveillance Program during the 2011/12-2014/15 academic years. The NCAA Injury Surveillance Program has been in existence since the early 1980s and have been assisting the NCAA in assessing sport-related safety in their sponsored programs. Participation in the NCAA Surveillance Program varied by sport and academic year. Data were collected by team athletic trainers that worked with these sport programs during the season.
We computed four measures of concussion incidence in a 13 different sports:
|Men’s sports||Women’s sports|
– Ice Hockey
– Field hockey
– Ice hockey
The four measures are described in the table below.
At what rate are concussions sustained during at-risk exposures?
Example: Across 10,000 NCAA football athlete-exposures, we expect to see 6-7 concussions
|One-season risk of concussion
What is the probability of an athlete obtaining a concussion in one season?
Example: In one season, we expect 1 in 20 NCAA football players to have a concussion?
|Average # concussion per team per season
How many concussions does a team sustain in one season?
Example: In one season, we expect 5-6 concussions within a NCAA football team
|% teams with a concussion
How many teams have a concussion occur within a season?
Example: In one season, we expect 80.6% of all NCAA football teams to have at least one concussion
The computations for the measures included alongside rates are included below:
What we found is that despite some variation in the rank-order of included sports, full contact sports such as wrestling, football, and ice hockey consistently generated the highest incidence of concussion.
Despite the similar rankings across sports, we believe that squad size may serve as a confounder, particularly in football. Furthermore, such measures can be biased when comparing incidence across teams (or sports) that vary greatly by the number of athletic sessions per season. Thus, it is important for readers to understand the strengths and limitations of measures of injury incidence utilized by various researchers.
How do these findings impact the public?
Although injury rates are the most preferred method of gauging injury incidence in academia, they may not be intuitive to non-scientists, including members of sports organizations concerned about the incidence of injury among their players. To help parents, coaches, and athletes, and to drive the development of data-driven, evidence-based policy and rule changes, we need to ensure that we are providing our findings in an easily understood manner.
This research presents a collection of “alternative facts” that still utilize the data collected by athletic trainers in a valid manner, but may be easier to interpret and disseminate to stakeholders. Better yet, these measures are applicable to other injuries and settings. This more diverse “toolbox” of measures, in combination with traditional athlete-based rates, may help sports organizations better identify specific athletes at risk for injury.