So, I have an interesting question for #epitwitter and science fans
Is this study actually a randomized controlled trial (RCT)?
https://pubmed.ncbi.nlm.nih.gov/33150385/ pic.twitter.com/kaqkerQHvd
You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
Splitting up your big trial into two papers is a bit odd, but with the perverse incentives of academia (2 pubs is better than 1!) it makes sense so all good Except then we get to these paragraphspic.twitter.com/cIN5ECynIj
Looking at the final sample size in the paper, it turns out to be <50% of the final sample. So we've got a per-protocol analysis of <50% of the total sample that can't be done ITT because the data is missing 

pic.twitter.com/nzFGbv9xre
When did the authors decide to conduct this secondary analysis? Let's check the study registry!pic.twitter.com/8YhWy5WOBN
Looking at the history... Study started recruiting - Jan 2013 Study finished recruiting - Jan 2019 Secondary outcomes added to register - June 2018pic.twitter.com/gA964fhoWI
In other words, the study went for 5 1/2 years before adding these secondary outcomes, which might go a way towards explaining why >50% of the participants have missing data
Back to the original question - given that this is a per-protocol analysis of a subset of the study, with non-random selection on an outcome selected 5 years after the study began, is it fair to publish as an RCT?
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.