I get asked a lot on @justsaysrisks how the risks are calculated
There are a couple of wayshttps://twitter.com/SiFTW/status/1191300598123118592 …
-
-
For example, sometimes the relative risk is based on a meta-analysis of case-control studies Redoing the work required to calculate that risk if the study doesn't include it takes DAYS
Show this thread -
In these cases, I use the population level risk of the event to calculate an estimated absolute risk increase based on the figures in the study
Show this thread -
So, say they found a relative increase of 20% and the population prevalence/incidence (depending on study) is 0.1%, the absolute risk increase would be around 0.02% (this is an estimate so I use ~ to indicate uncertainty)
Show this thread -
In the particular case from the top of this thread, the absolute risk is readily calculable from publicly available CDC mortality data, which is what the insurance company in the article used to get their relative figure in the first place, expressed as a % chance
Show this thread
End of conversation
New conversation -
-
-
Is it okay to use raw figures given the relative risks / hazard ratios are always adjusted?
-
It'd be better too use adjusted AR but I'd need the raw figures for that
that's why the AR is always something of an estimate - Show replies
New conversation -
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.