Epi in the News

Hi everyone,

Thank you for sharing this article. A few thoughts on your questions:

1. Without a proper definition (and diagnostics) it is impossible to know which athletes are affected by CTE and which are not. Without definition and diagnostics it is not possible to say anything about the prevalence of CTE. And that knowledge is necessary to know the burden of CTE in athletes.

2. Strict inclusion criteria for a review or meta-analysis will help make that the result that will be found has strong evidence. At the same time strict criteria will make it hard to find suitable studies that can be included. Which might be a problem especially regarding to CTE because for example the clear definition / diagnostics is still missing. So which studies with what kind of patients should be in- or excluded?

3. I think cross sectional studies could be included as well. Include athletes that have been exposed (concussion) and also have the outcome (CTE). If information is available about the concussions the athletes had and/or possible other causes of CTE.

4. I guess an RCT would be best. But it is ethically not appropriate to expose athletes to concussions. Or let that happen without them knowing (double-blind). Which makes it hard to design any trail with concussion as the experiment/exposure.
I guess collecting data over a long time with lots of athletes would help in this. But I am not sure if it is possible without a RCT to determine a causal relationship between exposure (concussions) and outcome (CTE).

Looking forward to hear from others.

Kind regards,
Corine Knulst-Verlaan

Hello everybody..
Another very interesting read! I personally have a problem with playing any kind of sport because I am too scared :-)). But I follow cricket with all my heart and so am absolutely terrified of injuries on field. Phil Huges traumatized us fans as well...

Coming to the questions asked:

1) A case definition is a requirement for any condition- the diagnosis is based on this, and of course calculation of prevalence is based on this. If there is no clear definition, its becomes difficult to evaluate the true quantity of the problem- as a majority will not be diagnosed. If the true prevalence is not known, then the actual burden will not be known. It would be difficult to establish any cause effect relationship or find associations with exposures. Hence, it would be difficult to recommend any changes in the practice to reduce the prevalence or in this case, safeguard the health of the players.

2) The pros would be the internal validity. Having very strict inclusion criteria would be able to reduce the noise, and bring together very similar studies that can be consolidated easily and with much less chance for bias. This would mean that the results would be more reliable for the cases that satisfy those particular inclusion criteria.
The main problem with this, especially when you have a condition that does not have a clear definition, could be that the major bulk of the evidence would be ignored. All the conclusions would be based on the small number of studies that satisfy the inclusion criteria. And thats not how life works. The results from such an analysis would not be generalizable to the the majority and hence loses its external validity.

3) The have de-prioritised cross sectional studies, and also case- control studies. The issue with case-control might be that the clinical syndrome has not been fully defined, but ppl are being diagnosed with possible CTE. Case reports are also valuable evidence especially in new diseases where the diagnostic criteria and risk factors are still being explored.

4) I dont believe that an RCT is required to prove a cause effect relationship. There are many fields where the evidence is based on long term cohort studies, because controlling the exposure would be unethical (for example in pediatrics and obstetrics). A good evidence would be a well designed prospective comparative cohort study- with appropriate matching or adjustment for confounding done. Nowadays, there are many ways to account for the bias in observational studies- like propensity score matching, regression models etc. A well designed prospective observational study might even be better than an RCT to determine a cause effect relationship, because first and foremost observational studies are real life studies, when compared to the controlled environment of an RCT.
Another option would be a quasi-experimental study. For example- we can look at 2 teams with players who have repeated exposures to head trauma. An intervention, like extended breaks or skipping matches following head injury, can be offered to one team.. The two teams can then be followed for a period of time to look for outcomes.
The main problem here is that CTE would be expected to develop after quite a number of years after the injuries- which would make these studies really really long. All in all it would be challenging....

Fathima

I agree with Corine and Fathima.
I fear that the CISG selection of evidence and consequent findings are motivated by their wishing sport to continue as it is, and possibly by the money that is made by this.
They have ignored case control studies, for example showing that people with dementia after head injury have characteristic findings at post-mortem which are not seen in people with dementia from other causes. This evidence gives biological plausbility to the hypothesis that repeated trauma causes dementia by causing CTE. I think this evidence should be included in their review.
The decision about which evidence to include or not depends on the risk one is prepared to take with sports people's health. If one is not too concerned about this then a higher risk can be taken and weaker evidence can be ignored. If one is concerned for their health then a precautionary principle can be applied and weaker evidence can be considered.
In this case, in my opinion, strong biological evidence at post-mortem from case control studies is being ignored.
Don't knock yourself for being scared Fathima - perhaps you're just prudent!
Judith

Re #4 I'm always happy to share this important study, published in the BMJ: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC300808/

And....
https://pubmed.ncbi.nlm.nih.gov/30545967/