- My Library
- (Golden Oldie) Why screening tests to…
(Golden Oldie) Why screening tests to predict injury do not work – and probably never will…: a critical review
- Screening (or periodic health evaluations) remains important, but not to predict injuries.
- It is important to differentiate between injury prediction and risk estimation – they are not the same.
- Sport injuries are complex phenomena, and we must appreciate the volatility, uncertainty, complexity, and ambiguity when implementing strategies to mitigate the risk of injury.
BACKGROUND & OBJECTIVE
In order to create a targeted prevention programme it is essential to understand the risk factors and injury mechanisms that play a role in the occurrence of sports injuries (1,2).
It remains the elusive question that has been at the forefront of screening programmes globally – can a reliable and valid test to determine risk of injury and/or illness predict which athletes will develop an illness, or present with injury? And if identified, can we address any of these risk factors through a targeted intervention programme to effectively eliminate injuries?
This seminal paper addressed if and how a periodic health examination (PHE) to screen for risk factors for injury can be used to mitigate injury risk.
It is not suggested that screening be disregarded in establishing risk of injury, but rather refined in how we might approach our understanding of risk.
It is useful to compare screening in sports medicine to screening for general disease. While screening for disease such as breast cancer involves detecting established disease as early as possible, screening for injury risk usually involves using a performance test to detect impairments which predispose the individual to injury (e.g. hamstring weakness). When screening for disease, the outcome is dichotomous (healthy/sick; yes/no). When screening for sports injury, the outcome is usually continuous, therefore, the continuous variable must be translated to a dichotomous outcome (e.g. at risk of injury or not; yes/no).
When screening for disease, the objective is to initiate treatment as early as possible. In sports injury prevention, the objective is early intervention to minimize the risk factor before injury occurs (e.g. strength training for hamstring weakness). Risk factors can be modifiable and non-modifiable, and screening tests typically measure modifiable factors such as strength or control. However, it should be noted that non-modifiable factors (such as gender or previous injury history) can be used as well to target intervention measures to the subgroup thought to be at increased risk.
Developing a screening test
Two reasons are proposed for doing research on risk factors for injury:
To help understand why injuries happen
To identify who is at risk of injury
Typically, exploratory studies will have a cohort of athletes undergo a series of tests during preseason to identify potential risk factors and record who gets injured. If an association is identified between one or more factors and injury risk, it is often concluded that these can be used to predict who is at risk of injury. But that is not enough. When such results are used to predetermine cut-off criteria to separate athletes with high risk, the test fails to predict who will get injured. Regardless, efforts to then address a modifiable risk factor to reduce injury has failed.
Furthermore, we haven’t mentioned comparing an intervention based on a screening test in a randomized control trial. For a screening test to be relevant, it needs to capture the majority of athletes at higher risk of injury and be able to separate out athletes at low risk of injury to improve the efficacy of intervention programmes. We are yet to identify such a test, or even battery of tests.
Screening test properties
The ability of a test to predict injury is captured by determining its sensitivity (does the test capture all those with injury), specificity (does it capture only those with injury), positive predictive value (how many with a positive test are injured) and negative predictive value (how many with a negative test are not injured). In sport, sensitivity and specificity are inversely related; if you want to capture all injured players (100% sensitivity), specificity suffers (more uninjured athletes will be classified as having high risk). The critical question is where the cut-off value separating high-risk and low-risk groups should be set?
An Example (3)
The relationship between hamstring injury risk and various strength measures in 614 football players was explored over four seasons, with 190 of these players suffering a hamstring strain injury. Eccentric hamstring strength normalized to bodyweight at 60°/s was independently associated with injury risk (odds ratio 1.37 per 1 Nm/kg difference). However, as illustrated in figure 1, again there is substantial overlap between injured and uninjured players, which clearly illustrates that a screening test based on eccentric hamstring strength cannot be used to predict injury risk.
Even more appropriate statistical measures, such as receiver operating characteristic curve analyses revealed an area under the curve of only 0.56 (eccentric hamstring strength), where a value of 1.0 indicates perfect prediction and 0.5 indicates a truly useless test (one no better at identifying true positives than flipping a coin).
Even when interpreting categorical risk factors (such as previous history of injury, yes or no), results are still too variable to predict injury, and since an effect is observed in the group with no history of previous injury, we might still be tempted to provide an injury prevention programme to the entire team.
There may be a significant relationship between a specific test result and injury risk but this is insufficient to use the test to predict who is at risk of injury.
There have been few programmes that have attempted to follow these steps in determining a screening test or PHE that will help to identify athletes at high risk of injury, impose an intervention, and thereby reduce the number of injuries. New models of dynamic systems theory and machine learning may aid us in developing algorithms or clinical prediction rules that are useful in identifying those athletes at high risk of injury.
A renewed emphasis has been placed on the context for athletes, their environment, and other important factors that will influence the outcome. The complex nature of sports injuries needs to be acknowledged, and importantly, we need to understand what interventions are successful for an individual or a team, if at all.
Although prospective risk factor studies may demonstrate a highly significant relationship between certain factors and injury risk (therefore improving the understanding of potential causative factors), such tests are unlikely to be able to predict injury with sufficient accuracy. While predicting future injury risk through screening tests is unrealistic, a PHE or pre-participation examination (screening) can serve several other purposes (4):
A comprehensive assessment of the athlete’s current health status, and typically, it is the entry point for medical care of the athlete.
Build a therapeutic alliance (good relationship) between the medical team and the athlete.
Review medications and supplements to avoid inadvertent doping.
Establish a performance baseline for the athlete in the healthy state.
Medicolegal duties of care
It is not suggested that screening be disregarded in establishing risk of injury. Rather, in how we might approach our understanding of risk, and how regular monitoring of relevant factors can help to improve the methods we apply when interpreting results, to have a greater appreciation for the clinical context of the questions we aim to answer.
- Meeuwisse WH. Assessing causation in sport injury: a multifactorial model. Clin J Sport Med 1994;4:166–70.
- Bahr R, Krosshaug T. Understanding injury mechanisms: a key component of preventing injuries in sport. Br J Sports Med 2005;39:324–9.
- Van Dyk N, Bahr R, Whiteley R, Tol JL, Kumar BD, Hamilton B, Farooq A, Witvrouw E. Hamstring and quadriceps isokinetic strength deficits are weak risk factors for hamstring strain injuries: a 4-year cohort study. The American journal of sports medicine. 2016 Jul;44(7):1789-95.
- Ljungqvist A, Jenoure P, Engebretsen L, Alonso JM, Bahr R, Clough A, De Bondt G, Dvorak J, Maloley R, Matheson G, Meeuwisse W. The International Olympic Committee (IOC) Consensus Statement on periodic health evaluation of elite athletes March 2009. British journal of sports medicine. 2009 Sep 1;43(9):631-43.