Why do injury-prediction screening tests fail?

Major points

Nevertheless, screening (or routine health checks) are no longer useful in predicting injuries.
2. It’s critical to recognize that risk estimate and harm prediction are two distinct processes.
3. When adopting solutions to reduce the risk of injury, we must be aware of the volatility, unpredictability, complexity, and ambiguity of sport injuries.

CONTEXT AND OBJECTIVE

Understanding the risk variables and injury processes that contribute to the development of sports injuries is crucial for developing a tailored preventative approach (1,2). Can a reliable and accurate test to identify risk of injury and/or sickness predict which athletes will acquire an illness or present with an injury? This elusive issue continues to be at the forefront of screening programs across the world.
Can we successfully reduce injuries by addressing any of these risk factors via a tailored intervention program, assuming they are shown to exist? This ground-breaking study examined whether and how a periodic health examination (PHE) to check for injury risk variables might be utilized to reduce injury risk. Understanding the risk variables and injury processes that contribute to the development of sports injuries is crucial for developing a tailored preventative approach (1,2). Can a reliable and accurate test to identify risk of injury and/or sickness predict which athletes will acquire an illness or present with an injury? This elusive issue continues to be at the forefront of screening programs across the world. Can we successfully reduce injuries by addressing any of these risk factors via a tailored intervention program, assuming they are shown to exist? This ground-breaking study examined whether and how a periodic health examination (PHE) to check for injury risk variables might be utilized to reduce injury risk.

Methods

It is helpful to contrast sports medicine screening with general illness screening. Although early detection of existing illness is important in screening for diseases like breast cancer, screening for injury risk often entails utilizing a performance test to find impairments that make a person more likely to sustain an injury (like hamstring weakness). The result of a disease screening is binary (healthy/sick; yes/no). Given that the result of a sports injury screening is often continuous (e.g., at risk of injury or not; yes/no), the continuous variable must be converted to a dichotomous result.

The goal of illness screening is to begin therapy as soon as feasible. In order to avoid sports injuries, early intervention is desired to reduce the risk factor before an injury develops (for example, strength training for hamstring weakness). Screening tests often evaluate changeable variables like strength or control, which are both modifiable and non-modifiable risk factors. It should be highlighted, however, that non-modifiable characteristics (such gender or a history of prior injuries) may also be utilized to focus intervention measures to the subgroup deemed to be at higher risk.

It is proposed that screening be improved rather than ignored when determining harm risk on how we could go about approaching risk knowledge.

Result

There are two justifications put up for studying injury risk factors:

1) To provide light on the causes of injuries
2) To determine who is most at risk of harm
Normally, preseason testing is conducted on a cohort of athletes as part of exploratory research to discover possible risk factors and track who sustains injuries. It is often inferred that these may be used to predict who is at risk of harm if a relationship is found between one or more characteristics and injury risk. Yet that is insufficient. The test is unable to identify which athletes are at a high risk of injury when such findings are used to establish cut-off criteria. Nonetheless, attempts to then address a modifiable risk factor to minimize harm has failed.
In a randomized control experiment, we haven’t spoken about contrasting an intervention based on a screening test. A screening test must be able to identify the majority of athletes who are at greater risk of injury and be able to distinguish between athletes who are at low risk of injury in order to increase the effectiveness of intervention programs. Such a test, or perhaps a battery of tests, have not yet been identified.

Properties of screening tests

Determining a test’s sensitivity (does the test accurately anticipate harm) captures the capacity of the test to do so.

catch every person who has an injury), specificity (does it just capture those who have an injury), positive predictive value (how many people who test positive are wounded), and negative predictive value (how many people who test negative are not hurt). Sensitivity and specificity in sports are adversely correlated; for example, if you wish to identify every injured player (100% sensitivity), specificity decreases (more healthy athletes would be labeled as high risk). Where should the cutoff value be placed to differentiate between high-risk and low-risk groups?

A Case Study (3)
During the course of four seasons, 614 football players were examined to determine the association between hamstring injury risk and different strength metrics; hamstring strain injuries were sustained by 190 of these players. The probability of injury was independently correlated with eccentric hamstring strength adjusted to bodyweight at 60°/s (odds ratio 1.37 for 1 Nm/kg difference). A screening test based on eccentric hamstring strength cannot, however, be used to predict injury risk, as shown in figure 1 where there is once again a significant overlap between affected and uninjured athletes.
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 a value of 0.5 indicates a truly useless test (one that does no better than flip a coin to identify true positives). Notwithstanding how categorical risk variables are interpreted (such as previous history of injury, yes or no),The findings are still too varied to forecast damage, and since an effect is shown in the group with no history of past injury, we may be on to something.

You could still be tempted to provide an injury prevention program to the whole team.
A strong link may exist between a certain test result and injury risk, but this is inadequate to utilize the test to determine who is at risk of damage.

IMPLICATIONS FOR PRACTICE

Despite the fact that studies of prospective risk factors may show a strong correlation between certain characteristics and the probability of harm,

Such tests are unlikely to be able to predict harm with adequate precision (despite advances in our knowledge of probable causal elements).
While it is impractical to forecast future injury risk using screening tests, a PHE or preparticipation examination (screening) may still be used for a number of additional objectives (4):
1) A thorough evaluation of the athlete’s present state of health, which often includes

It serves as the starting point for the athlete’s medical treatment.
2) Establish a therapeutic partnership (good relationship)

between the athlete and the medical staff.
3) Examine vitamins and drugs to prevent unintentional doping.
4) Set a baseline of the athlete’s performance in a healthy condition.
5) Medical-legal care obligations
It is not recommended that screening be ignored when determining harm risk. Instead, we should consider how we may approach our awareness of risk and how frequent monitoring of pertinent aspects might assist to enhance the techniques we use when interpreting data and to have a better grasp of the clinical context of the questions we intend to answer.