Thread by @Dr2NisreenAlwan

1/ Systematic review of the prevalence of #LongCovid

Now out as a preprint (not peer reviewed yet). It was a LOT of work. So grateful to my co-authors for their monumental effort: @MirembeWoodrow @NidaZiauddeen Darren Greenwood @vlutje1 Rebecca Thomas, @AthenaAkrami Chales Carey

2/As expected we found a huge variation between studies (0-93%). Prevalence generally higher in studies with mostly hospitalised patients. Because of our search period, studies estimating prevalence post vaccination were not included but we know these generally show lower numbers

3/ The main message is that the way #LongCovid is defined & measured affect prevalence estimation. However, even with the most conservative estimates, given past and continuing #SARSCoV2 infection, the burden of chronic disease post infection on population health is substantial.

4/Despite large between-study heterogeneity the studies assessed as with low risk of bias
estimated prevalence between 3-37%. We used this tool to assess bias. We learned along the way so there's room for improvement. You're welcome to use it citing paper: medrxiv.org/content/10.110…

5/ Study differences were mainly around #LongCovid definition, required threshold of severity or impact on daily activities, study designs, sources of study samples, how initial infection was defined, number of assessed symptoms & method of
assessment. We need consensus on these!

6/ Needless to say, this review was challenging to synthesise. We did our best but if we were striving for perfection we could've gone on forever! There's a wealth of supplementary material & plots. Pls feel free to explore. I hope we can update following peer review too.

7/ Downlaod paper from here: medrxiv.org/content/10.110…

#CountLongCovid is complex but we must keep on trying. Reaching strict definitions & rules to count it can be challenging with a real risk of exclusion, particularly that there's still so much to learn. Let's keep talking!