FYI, when I took Drug Study Design during my PharmD coursework back in the early to mid 90's, the faculty was from McMaster University in Canada. Guyatt is the guru. There is a systematic approach to evaluating the base of evidence to inform practice.
As has been mentioned here, sometimes you don't have good RCTs to inform practice. You do your best with extant data. If you are a well-trained clinician, you use your experience, intuition, and things you learned from pedagogic teachings to fill in the gaps.
Now here's the thing. If you DO have good RCTs that point in one direction, but you want to bend towards a bias (wherever the bias comes from, ie cute sales rep), you deserve whatever malpractice suits and negativity headed your way...especially when your patients are harmed.
All this said, in the absence of effective, curative medicines, there are numerous strategies clinicians must employ. There's no question that politics has entered the arena on this approach, which is unfortunate because some clinicians are still trying to do their best with what they have in front of them.
Bottom line: multiple RCTs demontrated no benefit of HCQ, which outweighs the supportive evidence base, which consists of zero blinded, well-designed RCTs that I can find. If the scientific community can sort through everything learned thus far regarding HCQ, I have no issue with informing new strategies to be tested via adequately powered, well designed RCTs, although Anastasis make a good point about the competition between studies and the precious resource of how many eligible patients are proximal to qualified research sites.