Filing early beats arguing later. And ask for the reference number, always.

Three questions kill most of the confidence in a thread like this: sample size, blinding, who paid.

Sleep first. Bad night, bad day, more symptoms. Fix that before changing anything else.

Same shape here, different drug. The thing that helped was documenting the refusal rather than re-arguing it. They refer rather than write.

Ten percent cuts, of the current dose not the original. That is the arithmetic that keeps it survivable.

If it needs three other compounds to notice, you have not tested the first one.

Read the methods section. It is the only part that says whether the conclusion is allowed to exist.