In article ,
James Warren wrote:

Epidemiology can never be a substitute for a random controlled
study. Epidemiology studies cannot control for confounding.

And, of course, drug companies can easily be confounded especially
when it's profitable.

--
This space unintentionally left blank.