That's true for simple things. You don't need to know what the pharmacological mechanism of COX inhibitors is in order to prescribe Advil for a headache. But if you're a scientist trying to make a better Advil you probably need to know how it works.
Doctors routinely prescribe medications that have no randomized clinical trials supporting their use. In those cases, clinical experience replaces trial data; they "know" the drugs work because all the patients have effectively been trial subject over a span of decades.
Doctors routinely prescribe medications that have no randomized clinical trials supporting their use. In those cases, clinical experience replaces trial data; they "know" the drugs work because all the patients have effectively been trial subject over a span of decades.