-
-
Replying to @jamesheathers
Health Nerd Retweeted Andrew Althouse
I reckon no chance is a bit too strong. It's unlikely that you'd have patients who were so perfectly matched in such a small sample, but given what they've said about the model if they existed you'd expect to see pretty much thishttps://twitter.com/ADAlthousePhD/status/1280663206629978116?s=20 …
Health Nerd added,
Andrew Althouse @ADAlthousePhDReplying to @f2harrell @SteveJoffeJust to be clear - it is plausible (whether it was advisable is a different matter). As mentioned in other thread - since they say they matched “exactly” on the PS, it is plausible (even likely) that the pairs matched had exactly the same covariate values.2 replies 0 retweets 5 likes -
Replying to @GidMK
Sorry, what enormous basket of mothers was this selected from then???
1 reply 0 retweets 3 likes -
Replying to @jamesheathers
2,500 patients total, they took a subset of the HCQ prescribed people and matched to 190 of the non-HCQ people. Depending on the software/commands, that's what you'd expect to see if they matched exactly as they described
1 reply 0 retweets 0 likes -
Replying to @GidMK @jamesheathers
Also, the categories they matched on were pretty broad, and might include pretty much anything (i.e. "cardiovascular complications") so might just be a case of a ridiculous propensity score that was a waste of time rather than an error
1 reply 0 retweets 0 likes
That being said, I could probably reproduce that table in my dataset of 100,000 people, with only 2,500 it becomes pretty unlikely imo
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.