2020 was a year of relatively uninformative studies about the effects of COVID on the heart as I have previously lamented. 2021 has proved more informative. But where does this leave athletes following COVID-19 infection?
Puntmann et al’s study suggesting that asymptomatic COVID-19 infection could still carry significant cardiac risk raised the alarm for athletes last year. Rajpal et al’s study of 26 college athletes at Ohio State University further exacerbated these fears. These 2 studies’ findings were interpreted as meaning that cardiac MRI was needed to ensure athletes could safely return to sport. This was not a valid conclusion and Venkatash Murthy made sure everyone knew it.
Guidance for athletes was still needed though. One editorial described the issue: “The recommendations exemplify the major challenge facing these colleagues as they try to sculpt guidance from scant information and data. Expert opinion is far better than no opinion, but the recommendations call into stark light the usual conundrums involved in screening for low-prevalence findings.” I personally think no opinion is better than expert opinion without evidence but I realise people value having guidelines to lean on. It’s definitely easier than conveying the true uncertainty to patients.
Thank heavens for Moulson et al’s study published this week. Unlike Puntmann et al and Rajpal et al, this study allows us to draw conclusions about the prevalence of cardiac involvement in athletes following COVID-19 infection. The key to estimating prevalence is using the right sample – right in size (large enough) but more crucially, right in the way patients were selected.
How to choose a cohort for estimating prevalence? You need the right denominator. Examining cases of athletes with cardiac involvement after COVID-19 comes across as very persuasive but gives no prevalence information without the denominator, which in this case would be all athletes who had COVID-19. Moulson et al got the next best thing – all athletes who tested positive across National Collegiate Athletic Association (NCAA) institutions. The size and breadth of this sample across institutions – over 3,000 of over 19,000 young athletes tested across 42 institutions – add to the validity of the study’s findings. Unfortunately it’s not practical to do cardiac MRIs on all 3000 of these positive individuals. Instead, it was up to the institution whether to do a cardiac MRI.
A third of these 3000 positive individuals were asymptomatic. Three quarters had “triad” testing i.e. ECG, troponin and echo. 119 had a clinically indicated cardiac MRI and 198 underwent a primary screening cardiac MRI.
21 individuals were determined to have myocardial or pericardial SARS-CoV-2 involvement by cardiac MRI. This finding was 4 times more likely in the clinically indicated MRIs compared to the screening ones. These data really do not support screening by cardiac MRI.
Predictors of cardiac involvement were cardiopulmonary symptoms or at least one abnormal test in the “triad” testing. There is plenty of lovely robust data here to base guidelines on.
The abstract concludes: “SARS-CoV-2 infection among young competitive athletes is associated with a low prevalence of cardiac involvement and a low risk of clinical events in short term follow-up”. Indeed. What a beacon of a study in terms of design and reporting of results!
So can we all agree COVID-heart in athletes isn’t an entity? Or at the very least, that we shouldn’t worry about it?
There’s a lesson here. It isn’t better to rush out publications with significant limitations just because of the importance of the topic. Moulson et al’s study is the kind of thing we should have waited for before jumping to conclusions – proper evidence.